Takeoff Minimums: Part 91 vs. 121/135, the Standard 1-Mile Rule, RVR Tiers, and Takeoff Alternates
Updated: Sep 30
Takeoff is one of the most critical phases of flight, especially in reduced visibility or instrument meteorological conditions. To manage risk during this phase, aviation authorities establish takeoff minimum regulations — rules that govern when an aircraft may legally depart based on weather, visibility, aircraft capability, and operational context. But the rules are strange in a way that surprises new instrument pilots: for most general aviation flying, there ARE no legal takeoff minimums. You can depart in fog you can't see across. That legal freedom is real — and it's a trap, because the physics of an engine failure at 200 feet in zero-zero don't care which Part you operate under. Understanding what the regulations require, what the commercial operators must do, and how to apply their logic to your own decisions is essential for safe, legal, and professional operations.
This post covers takeoff minimums in practical depth: what they are and how they differ from landing minimums, the Part 91 reality (and its caveats), the standard minimums under 91.175(f), RVR and how it converts to visibility, the lower-than-standard tiers and what they require, non-standard published minimums and how you find them, takeoff alternates with the actual rule, the one case where ceiling matters, and the practical decision test every pilot should use.
Study this full length lesson (video, podcast, flashcards, and quiz) here: Full Length Lesson >
What Takeoff Minimums Are
Takeoff minimums define the lowest weather conditions under which an aircraft is permitted to BEGIN a takeoff.
They're typically based on:
Visibility (statute miles, or runway visual range in feet)
Ceiling (only in specific cases — below)
Runway lighting and markings (which enable lower visibility minimums)
Available departure procedures (ODPs, climb gradients)
Aircraft equipment and crew qualification
How they differ from landing minimums:
Landing minimums are always explicit (a DA/MDA and a visibility on every approach chart)
Takeoff minimums are NOT always stated — and for Part 91, often don't legally exist
Landing minimums include a ceiling/altitude component; takeoff minimums are almost entirely about VISIBILITY
Two very different regulatory structures
The core question they answer:
"Is it legal to start the takeoff roll in this weather?"
Not "is it safe" — that's a separate (and more important) question
The regulations set the floor; judgment sets the standard
Legal ≠ prudent
The Regulatory Framework
Where the rules live.
The governing regulations:
14 CFR 91.175(f) — the "standard takeoff minimums" rule (and it's in Part 91, but it binds commercial operators — below)
14 CFR Part 121 — air carriers (with 121.617 for takeoff alternates)
14 CFR Part 135 — commuter and on-demand operators
14 CFR 91.13 — careless or reckless operation (the Part 91 backstop)
The supporting publications:
The Terminal Procedures Publication (TPP) — the "Takeoff Minimums, (Obstacle) Departure Procedures, and Diverse Vector Area" section
The AIM — guidance and recommendations
Operations Specifications (OpSpecs) — each commercial operator's specific authorizations
The key structural point:
The takeoff minimums regulation (91.175(f)) lives in Part 91
But it applies "to operations conducted under Part 121, 125, 129, or 135"
Part 91 operators are NOT bound by it
A regulation in Part 91 that doesn't apply to Part 91 — a classic confusion
Takeoff Minimums Under Part 91: The Freedom and the Trap
The Part 91 reality, expanded.
No explicit takeoff minimums:
For most general aviation operations under Part 91, there are NO regulatory takeoff minimums
A Part 91 pilot may legally take off in ZERO visibility
Even when the weather is below the published LANDING minimums for that airport
Even in "zero-zero"
What the regulation actually says:
91.175(f) sets standard takeoff minimums — but only for Part 121/125/129/135
No other regulation imposes a visibility requirement for Part 91 IFR takeoff
Hence: legally, none
The letter of the law
The caveats that remain:
Obstacle clearance is still your responsibility (fly the ODP)
91.13 (careless or reckless): a zero-zero takeoff that goes wrong will be examined under this rule — "legal" doesn't protect you from a careless/reckless finding after an accident
Aircraft capability: you still need to be able to fly the departure
The freedom isn't unlimited
Why the trap is real — what zero-zero actually means:
You cannot return and land. If anything goes wrong in the first minutes — engine roughness, a door pops, smoke, an instrument failure — the departure airport is below landing minimums. You're committed to flying somewhere else in IMC with a problem.
You're on instruments from liftoff. No visual transition; the gyros must be trusted the instant the wheels leave, when you're at max workload
Limited climb performance at heavy weight/high density altitude compounds it
Spatial disorientation risk peaks in the initial climb with no visual reference
The margins are thinnest exactly when the freedom is greatest
The engine-failure scenario:
Engine failure at 300 feet in zero-zero, in a single
No runway to return to, no visibility to pick a field
The odds are poor
This is why the freedom exists on paper but shouldn't be exercised in practice
Personal minimums:
Most experienced pilots adopt personal takeoff minimums well above zero
A common approach: don't take off unless you could come back and land (below)
Others use the commercial standard (1 SM) as a personal floor
Legal freedom, self-imposed discipline
Standard Takeoff Minimums: The 91.175(f) Rule
The actual numbers the original lists but doesn't tie to the rule.
The standard minimums (91.175(f)):
1 statute mile visibility for aircraft with 1 or 2 engines
½ statute mile visibility for aircraft with 3 or more engines
These apply when no non-standard minimums are published for the runway
The "standard" everyone references
Who they bind:
Part 121, 125, 129, and 135 operators
NOT Part 91 (though Part 91 pilots are encouraged to use them)
The commercial floor
Why more engines = lower minimum:
A 3- or 4-engine aircraft losing one engine retains a large performance margin
A twin losing one engine loses half its power (and may have marginal climb)
Better engine-out performance justifies lower visibility
Redundancy earns the lower minimum
The "std" notation in the TPP:
When the TPP takeoff minimums section says a runway is "std" (standard), it means the 91.175(f) minimums apply (1 SM / ½ SM)
Often written with a climb gradient: "Rwy 27, std. w/min. climb of 300' per NM to 3000"
"Std" is shorthand for the 91.175(f) values
Know what it points to
Non-standard overrides standard:
If the TPP publishes a HIGHER minimum for a runway (e.g., "Rwy 9, 300-1" or "Rwy 18, 1 SM w/climb of 400' per NM"), that replaces the standard
Non-standard minimums are always MORE restrictive than standard (or add a gradient)
Published minimums win
Read the TPP
Runway Visual Range (RVR) and How It Converts
RVR, with the conversion table the original omits.
What RVR is:
A measurement of how far down the runway a pilot can see, reported in FEET
Measured by transmissometers or forward-scatter sensors alongside the runway
Runway-specific and instrument-measured (not a human estimate)
The standard for low-visibility operations
Why RVR is better than reported visibility:
More precise (feet, not fractions of a mile)
Runway-specific (fog can vary across an airport)
Updated continuously
Critical for low-visibility takeoffs and landings
The RVR-to-visibility conversion (from the FAA table):
RVR (feet) | Visibility (statute miles) |
1,600 | ¼ SM |
2,400 | ½ SM |
3,200 | 5/8 SM |
4,000 | ¾ SM |
4,500 | 7/8 SM |
5,000 | 1 SM |
6,000 | 1¼ SM |
Why the table matters:
Standard minimums are stated in SM (1 SM, ½ SM)
RVR is reported in feet
You convert: 1 SM = RVR 5000; ½ SM = RVR 2400
When RVR is reported, it CONTROLS (over the reported visibility)
The RVR-controlling rule:
If RVR is reported for the runway, use it (not the tower-reported prevailing visibility)
RVR is more accurate and runway-specific
If RVR isn't available, use reported visibility
RVR takes precedence
The multiple-sensor rules (for low-visibility takeoffs):
Long runways may have touchdown, mid-point, and rollout RVR sensors
For lower-than-standard takeoffs, specific sensors are "controlling" (typically the touchdown RVR controls; mid and rollout become controlling as minimums get lower)
Specified in the operator's OpSpecs
The lower the minimum, the more sensors must report
Lower-Than-Standard Takeoff Minimums: The Tiers
The commercial-operator world the original gestures at — with the actual tiers.
The concept:
Commercial operators can be AUTHORIZED (via OpSpecs) to take off BELOW the standard 1 SM / ½ SM
Each lower tier requires MORE runway lighting/equipment and MORE crew training
The equipment substitutes for the missing visibility
Authorization, not entitlement
The typical tiers (via OpSpecs — values vary by operator):
RVR 1600 (¼ SM): requires at least one of: operative centerline lights (CL), runway centerline markings (RCLM), or adequate visual reference (e.g., high-intensity runway lights)
RVR 1200: requires operative centerline lights AND typically two RVR sensors reporting
RVR 1000: requires centerline lights plus (in many OpSpecs) a HUD or takeoff guidance system, and multiple RVR sensors
RVR 500/300: the lowest, requiring a Head-Up Display with takeoff guidance, centerline lights, and all RVR sensors
Each step down adds requirements
The logic of the tiers:
As visibility drops, the pilot needs MORE cues to stay on the centerline during the takeoff roll and an engine-failure abort
Centerline lights and markings provide those cues
At the lowest levels, a HUD provides guidance the eye can't
Equipment replaces visibility
"Adequate visual reference":
A term in the OpSpecs meaning the runway has enough lighting/marking for the pilot to maintain directional control
Specifically defined (e.g., high-intensity runway edge lights, centerline lights, or centerline markings)
The minimum cue set for each tier
A defined standard, not a judgment call
Why Part 91 pilots should understand this:
It shows how seriously the professional world treats low-visibility takeoff
Airlines with two pilots, a HUD, and centerline lights still won't go below RVR 500
A Part 91 single-pilot in a light aircraft is legally free to go in RVR 0
The contrast should inform your personal minimums
Published Takeoff Minimums and Departure Procedures (TPP)
Where the non-standard minimums live, with the finding mechanism.
The TPP section:
"Takeoff Minimums, (Obstacle) Departure Procedures, and Diverse Vector Area"
At the front of each TPP volume, listed alphabetically by airport
Contains, per airport and runway: standard or non-standard takeoff minimums, ODPs, and DVAs
One place for all of it
How you know to look — the Trouble T:
The approach charts for the airport carry a black triangle with a white "T" if non-standard takeoff minimums OR an ODP exist
The T is your pointer to the TPP section
No T = standard minimums and a diverse departure
Check every departure airport
When non-standard minimums are published:
Terrain or obstacles penetrate the obstacle identification surface (requiring higher visibility so you can see and avoid, OR a climb gradient, OR a specific routing)
Runway lighting/marking is inadequate for lower visibility
Standard obstacle clearance can't be guaranteed on a straight-out climb
The airport has a problem the standard doesn't cover
What non-standard minimums may require:
Higher visibility (e.g., "Rwy 27, 1 SM" or "300-1" — which includes a ceiling)
A specific climb gradient ("std. w/min. climb of 350' per NM to 5000")
Use of a particular runway ("Rwy 18, NA" — not authorized for IFR departure)
A visual climb over the airport (VCOA — below)
The "NA" runway:
Some runways are marked "NA" for IFR takeoff (obstacles too severe, no procedure possible)
Legal for VFR departure, not IFR
Use another runway
Check for it
The Part 91 relationship to published minimums:
Part 91 isn't legally bound by the published visibility minimums
But the ODP and climb gradient published alongside them are still your obstacle-clearance plan
And the published visibility often reflects "you need to see the obstacle to avoid it"
Treat the published minimums as your working minimums
Visibility vs. Ceiling: When Ceiling Actually Matters
The core concept, with the exception.
The general rule:
Visibility is controlling for takeoff
Ceiling is NOT typically limiting — you can legally depart with a zero ceiling as long as visibility (if applicable) is met
Because takeoff is about seeing the runway and maintaining directional control, not seeing the ground from altitude
Different from landing (where ceiling/DA matters enormously)
Why ceiling doesn't matter for a normal takeoff:
Once airborne, you're on instruments regardless
A 100-foot ceiling and a 1,000-foot ceiling produce the same instrument departure
The visibility governs the ground portion (the roll and the abort)
Ceiling is irrelevant to the physics of the roll
The exception — when a ceiling IS published:
Some non-standard minimums include a ceiling: e.g., "Rwy 9, 300-1" (300-foot ceiling, 1 SM visibility)
This happens when the departure requires you to SEE AND AVOID an obstacle visually after liftoff, so you must have a ceiling high enough to see it
The ceiling ensures visual obstacle avoidance is possible
The "see and avoid" departure
The VCOA exception:
A Visual Climb Over Airport requires published ceiling AND visibility minimums (e.g., "VCOA: 2,000-3")
Because you're climbing visually over the airport, you need to see the airport throughout the climb
The one departure type where ceiling is central
(See our ODP/SID post for the VCOA procedure)
The takeaway:
Normal IFR takeoff: visibility controls, ceiling doesn't
Published ceiling minimums: only when the departure depends on visual obstacle avoidance (including VCOA)
Read the published minimums for a ceiling value
Know which situation you're in
Takeoff Alternates: The Rule With Numbers
The commercial requirement, with the regulation the original doesn't cite.
The concept:
A takeoff alternate is an airport you could divert to if you had to return shortly after departure but the DEPARTURE airport's weather is below LANDING minimums
Answers "where do I go if something goes wrong right after takeoff?"
Required for commercial operators in specific conditions
The safety net the zero-zero Part 91 pilot doesn't have
When it's required (Part 121, 121.617; Part 135 similar):
When the weather at the departure airport is BELOW the LANDING minimums for that airport (for the operator's approaches)
I.e., you can take off (visibility meets takeoff minimums) but couldn't land back (weather is below the approach minimums)
The "can take off but can't come back" gap
The trigger
The distance/time limits (121.617):
For 2-engine aircraft: the takeoff alternate must be within 1 hour of flying time at normal cruise speed in still air with one engine inoperative
For 3 or more engines: within 2 hours under the same conditions
Measured from the departure airport
The engine-out reach
What the alternate must offer:
Weather at or above the operator's landing minimums for that airport
A usable approach for the aircraft/crew
Within the time limit
A real place to land
Why it exists:
Protects against an immediate post-departure emergency when returning isn't possible
The commercial world refuses to launch into a "no place to land" situation
A structural safety requirement
Built into every low-weather departure
The Part 91 lesson:
Part 91 has NO takeoff alternate requirement
But the LOGIC is the best personal-minimums tool available (next section)
Borrow the professionals' thinking
Apply it voluntarily
The "Can I Get Back In?" Test
The practical decision tool, expanded from the takeoff-alternate logic.
The question:
Before any low-weather IFR departure, ask: "If something goes wrong in the first five minutes, can I get back in — here or somewhere close?"
If the departure airport is below landing minimums and there's no nearby airport above minimums, the answer is NO
That's a "no-go" for a prudent Part 91 pilot, regardless of legal freedom
The single most useful takeoff-minimums habit
How to apply it:
Check the departure airport's weather against its LANDING minimums (the approach you'd fly to return)
If above landing minimums: you can return — the risk is manageable
If below landing minimums: identify a nearby airport (within a few minutes' flight, given your aircraft) that IS above minimums
If one exists and you can reach it on one engine (or with your problem): acceptable, with a plan
If none exists: don't go
The single-engine reality:
In a single, an engine failure after a zero-zero takeoff has no good outcome
"Can I get back in?" for a single is really "can I get to a runway at all?"
Many single-engine IFR pilots set a floor of "departure airport at or above landing minimums"
The honest minimum for singles
The twin reality:
A twin can (in theory) return on one engine
But only if the departure airport (or a close alternate) is above landing minimums
The takeoff-alternate logic applies directly
Plan the return before the departure
Why this beats a fixed visibility number:
A fixed personal minimum (e.g., "1 SM") doesn't account for whether you can return
"Can I get back in?" ties the decision to the actual escape option
Weather that's legal AND allows a return is the real standard
Outcome-based, not number-based
Obstacle Clearance and Takeoff Minimums: Linked but Separate
The relationship, sharpened.
Two different problems:
Takeoff minimums address VISIBILITY — can you see enough to maintain directional control on the roll and abort safely?
Obstacle Departure Procedures address TERRAIN and OBSTACLES — once airborne, can you clear what's out there?
Related (both in the same TPP section) but distinct
Both must be satisfied
Where they intersect:
Some non-standard takeoff minimums exist BECAUSE of obstacles (higher visibility so you can see and avoid them)
Some ODPs include a climb gradient that's the real limiting factor (not the visibility)
The TPP entry often combines them: "Rwy 27, std. w/min. climb of 350' per NM to 4000, or 1100-3 for climb in visual conditions"
Read the whole entry
Even if visibility is legal:
The aircraft must still meet the required climb gradient (convert ft/NM to fpm at your groundspeed)
Multi-engine pilots must plan for the engine-out case (can you meet the gradient on one engine? Usually not — hence the takeoff alternate and the "get back in" logic)
Legal visibility with an unmakeable gradient is still a no-go
Two gates, both must open
(For the diverse departure standard, the 40:1 surface, climb gradient math, and VCOA, see our ODP vs. SID post.)
Pilot Responsibility and Risk Management
The judgment layer, expanded.
Regardless of the regulations, the PIC must consider:
Aircraft performance (climb gradient at actual weight/density altitude/temperature)
Density altitude (hot, high, humid kills climb performance)
Weight and balance (heavy = poor climb)
Crew proficiency (single-pilot IFR from liftoff is demanding — are you current and sharp?)
Emergency options after liftoff (the "can I get back in?" test)
Equipment reliability (a vacuum system you'd be trusting from the moment the wheels leave)
The single-pilot factor:
Airlines require two pilots, a HUD, and centerline lights before going below RVR 500
A Part 91 single-pilot has none of that support
Your personal minimums should reflect your resources, not the airline's
Honest self-assessment
The "legal does not mean safe" principle:
The regulations define the floor for commercial operators and NO floor for Part 91
That absence is not permission; it's a transfer of responsibility to you
Set personal minimums that account for your aircraft, your experience, and your escape options
The professional standard is self-imposed
Common Training and Checkride Questions — With Answers
"Are there takeoff minimums under Part 91?"
No — legally, Part 91 has no takeoff minimums; you can depart in zero-zero. But 91.13 (careless/reckless) still applies, obstacle clearance is still your responsibility, and the AIM strongly encourages using the published minimums. Prudent pilots set personal minimums (often: departure airport at or above landing minimums).
"What are the standard takeoff minimums?"
Under 91.175(f): 1 SM visibility for aircraft with 1 or 2 engines; ½ SM for 3 or more engines. They bind Part 121/125/129/135, not Part 91.
"When are non-standard takeoff minimums published?"
When obstacles penetrate the obstacle identification surface, when runway lighting/marking is inadequate for lower visibility, or when standard obstacle clearance can't be guaranteed. They're flagged by the Trouble T on the approach charts and listed in the TPP takeoff minimums section.
"What's the difference between takeoff minimums and ODPs?"
Takeoff minimums address visibility (the roll and abort); ODPs address obstacle clearance after liftoff (routing and climb gradient). Both are in the same TPP section and both must be satisfied.
"When is a takeoff alternate required?"
For Part 121 (121.617) and similar Part 135 rules: when the departure airport's weather is below the landing minimums for that airport. It must be within 1 hour one-engine-inoperative for twins, 2 hours for 3+ engines. Part 91 has no requirement — but the logic is the best personal-minimums tool.
"Does ceiling matter for takeoff?"
Generally no — visibility controls, and you can depart with a zero ceiling. Ceiling matters only when a non-standard minimum publishes one (a see-and-avoid departure) or for a VCOA.
"What does RVR 2400 equal in visibility?"
½ SM. (RVR 5000 = 1 SM; RVR 1600 = ¼ SM.) When RVR is reported, it controls over reported visibility.
Conclusion
Takeoff minimum regulations vary significantly by type of operation, but the underlying goal is always the same: manage risk during the most demanding phase of flight. Part 91 pilots have complete legal flexibility — no takeoff minimums at all — but that freedom is a transfer of responsibility, not a safety margin. Commercial operators must comply with strict published minimums: the 91.175(f) standard of 1 SM (or ½ SM for 3+ engines), lower-than-standard RVR tiers that demand centerline lights and HUDs as visibility drops, and takeoff alternates within one-engine-out reach whenever the departure airport is below landing minimums.
The professional world's structure offers the Part 91 pilot the best possible decision tool: before any low-weather departure, ask whether you could get back in. If the departure airport is below landing minimums and nothing nearby is above them, the airlines wouldn't go — and neither should you. All pilots must ensure obstacle clearance and climb performance regardless of visibility. A professional pilot understands not just what is legal, but what is prudent.
On the Written Test and Checkride
Takeoff minimums appear on tests and checkride orals (instrument and commercial especially). The most commonly tested topics:
Part 91 has no takeoff minimums (and the 91.13 caveat)
Standard takeoff minimums under 91.175(f) (1 SM / ½ SM) and who they bind
Visibility controls; ceiling generally doesn't
RVR and its conversion to visibility
Non-standard minimums (when published, the Trouble T, the TPP section)
Takeoff alternates (when required, the 1-hour/2-hour rule)
Takeoff minimums vs. ODPs
Quick Reference
Takeoff Minimums:
The lowest weather to legally BEGIN a takeoff
Almost entirely about VISIBILITY (ceiling only in special cases)
Different from landing minimums (which are always explicit)
Part 91:
NO legal takeoff minimums (zero-zero is legal)
Caveats: obstacle clearance (fly the ODP), 91.13 careless/reckless, aircraft capability
The trap: can't return, on instruments from liftoff, engine-out has no good outcome
AIM strongly encourages using the published minimums
Standard Takeoff Minimums (91.175(f)):
1 SM visibility: 1-2 engines
½ SM visibility: 3+ engines
Binds Part 121/125/129/135 — NOT Part 91
"std" in the TPP = these values
Non-standard published minimums override standard
RVR Conversion:
RVR | Visibility |
1,600 | ¼ SM |
2,400 | ½ SM |
4,000 | ¾ SM |
5,000 | 1 SM |
RVR controls when reported (runway-specific, instrument-measured)
Lower-Than-Standard Tiers (OpSpecs):
RVR 1600: centerline lights, markings, or adequate visual reference
RVR 1200: centerline lights + multiple RVR sensors
RVR 1000: + HUD/takeoff guidance (typically)
RVR 500/300: HUD with takeoff guidance + all sensors
Each lower tier = more equipment + more training
Non-Standard Minimums (TPP):
Flagged by the Trouble T on approach charts
Published when obstacles penetrate the surface, lighting is inadequate, or standard clearance can't be guaranteed
May require: higher visibility, a ceiling (e.g., "300-1"), a climb gradient, a specific runway, or a VCOA
"NA" = runway not authorized for IFR takeoff
Visibility vs. Ceiling:
Visibility CONTROLS; ceiling generally doesn't (zero ceiling is legal)
Ceiling matters only for see-and-avoid departures ("300-1") and VCOA
Takeoff Alternate (121.617):
Required when departure weather is BELOW LANDING minimums
Within 1 hour one-engine-inoperative (2 engines)
Within 2 hours (3+ engines)
Must be at/above landing minimums
Part 91: not required — but use the logic
The "Can I Get Back In?" Test:
Is the departure airport at/above its landing minimums?
If not, is a nearby airport above minimums and reachable with your problem?
If neither: don't go
Singles: departure airport at/above landing minimums is the honest floor
Takeoff Minimums vs. ODPs:
Minimums = visibility (the roll and abort)
ODP = obstacle clearance after liftoff (routing, climb gradient)
Same TPP section; both must be satisfied
Legal visibility + unmakeable gradient = still no-go
Key Principle:
Part 91 has no takeoff minimums — that's a transfer of responsibility, not a margin. The commercial standard (91.175(f)) is 1 SM for 1-2 engines, ½ SM for 3+, with lower RVR tiers requiring centerline lights and HUDs, and a takeoff alternate within one-engine-out reach whenever the departure airport is below landing minimums. Borrow that logic: before any low-weather departure, ask "can I get back in?" — and if the answer is no, the professionals wouldn't go either.
Study Full Aviation Courses:
wifiCFI's full suite of aviation courses has everything you need to go from brand new to flight instructor and airline pilot! Check out any of the courses below for free:
Study Courses:
Checkride Lesson Plans:
Teaching Courses:
Author: Nathan Hodell
CFI, CFII, MEI, ATP, Creator and CEO
Nathan is an aviation enthusiast with thousands of hours of flying and dual instruction over the past 15+ years. Through his aviation career he has been able to earn his ATP, fly as an airline pilot, own/operate flight schools, and create and host wifiCFI.