Halfway down the runway, the decision is already made. Either the aeroplane is accelerating the way the handbook said it would, or it is not, and in the second case the only variable left is how much runway remains to stop in. The 70/50 rule exists to force that decision to happen at the midpoint, while stopping is still possible, rather than at the far end, where it is not.

The rule is simple to state. If the airspeed has not reached 70 percent of the liftoff speed by the time the aircraft has used 50 percent of the runway, abort. It appears in the FAA’s Aeronautical Information Manual, in FAA safety literature, in mountain-flying courses and in almost every discussion of high-density-altitude departures. It is also, according to one of the more mathematically literate voices in general aviation, dangerous when misunderstood.

Both things are true, and the reason is arithmetic. This is what the rule says, where the 70 percent comes from, why the assumptions behind it fail in exactly the conditions that make it necessary, and how to use it so that it works for you rather than against you.

What the rule says, exactly

The FAA’s wording is worth reading in full, because most retellings sand off the caveats. Section 7-6-8 of the Aeronautical Information Manual, which deals with runway half-way signs at unimproved airports, states that assuming the runway length is appropriate for the takeoff, typical acceleration should allow the aeroplane to reach 70 percent of liftoff airspeed by the midpoint of the runway, and that if it does not, the takeoff should be aborted because it may not be possible to lift off in the remaining runway.

The same section then lists the conditions under which the rule is unreliable. Airspeed indicators in small aircraft are not required to be accurate below stalling speed and may be unusable at 70 percent of liftoff speed. The rule applies only to the runway needed to lift off; if there are obstacles on the climb path, additional distance is required to accelerate to best-angle-of-climb speed and clear them. And the technique does not account for upslope or tailwind. The FAA also notes that no standard exists for what a half-way sign looks like.

The FAA Safety Team (FAASTeam) restates the rule in a slightly different currency. Rather than 50 percent of the runway, it uses 50 percent of the calculated takeoff distance from the pilot’s operating handbook, marked by a landmark chosen before the roll begins, and it adds a second version for obstructed departures: 70 percent of rotation speed by 30 percent of the available distance. That distinction, runway versus computed distance, turns out to be the whole argument.

Where 70 percent comes from

Under constant acceleration from a standing start, distance is proportional to the square of speed. Written out: s = v² / 2a. Set the distance to half the liftoff distance and the speed at that point is the square root of one half times the liftoff speed, which is 0.707. Round it and you have 70 percent. That is the entire derivation.

Sparky Imeson, whose mountain-flying courses did more than anyone to spread the rule through the American backcountry community, generalised it: ten times the square root of the percentage of liftoff distance used equals the percentage of liftoff speed you should have. At 50 percent of the distance, the square root is 7.07, so 70.7 percent of the speed. At 25 percent of the distance, 50 percent of the speed. He also noted, with characteristic economy, why the check has to happen early.

Nobody has traced the rule to a single originator. Catherine Cavagnaro, a mathematics professor, designated pilot examiner and AOPA Pilot columnist, wrote in 2023 that she had spent years trying to find its source without success. One account, from a pilot writing in Air Facts Journal, links it to US Air Force acceleration-check charts used on early jets with very long takeoff rolls, where speed was checked at a fixed distance. That lineage is plausible but undocumented, and should be treated as such.

Where the arithmetic breaks

The derivation assumes constant acceleration. Real aircraft do not deliver it. Aerodynamic drag rises with the square of speed, propeller thrust falls as speed increases, and rolling resistance on a soft surface is not zero. The result is that the second half of a takeoff roll takes longer, and covers more ground, than the first half of the model predicts. An aircraft that just scrapes 70 percent at the midpoint is behind the curve, not on it.

Cavagnaro made the point sharply in AOPA Pilot. A takeoff that barely produces a go decision at the halfway point, she wrote, implies wheels leaving the ground in the last foot of runway with no margin for error. Her conclusion was that the rule is extracted from the most basic model of a takeoff, one that ignores air resistance entirely, and that achieving rotation speed by the end of the runway can require significantly more than 70 percent at the halfway mark.

Larry Brown, a retired US Air Force F-15 pilot writing for AOPA, attacked the rule from the other side: on a 10,000 ft runway in a light aircraft with a 60-knot liftoff speed, would you really want to be 5,000 ft down the runway at 43 knots and think things were fine? Applied to the runway rather than to the computed distance, the rule permits absurd acceleration on long runways and offers almost no margin on short ones. That is why the FAA Safety Team version anchors the check to the handbook figure instead.

The instructor community has responded with stricter variants. Some teach 75 percent at 50 percent. The FAA’s own obstacle version demands 70 percent by 30 percent. Cavagnaro’s position is that any rule short of reaching liftoff speed itself at a pre-planned point is a false comfort, and AOPA’s James Rush Manley recommends planning to lift off within 70 percent of the available runway with the abort point never past the halfway mark. Even Air Facts Journal, after publishing an explainer in 2025, appended a note that readers and experts had called the rule controversial.

Density altitude: the reason the rule exists at all

At sea level on a cool morning, a light aircraft accelerates so briskly that the midpoint check is a formality. The rule earns its keep at high, hot airfields, where a normally aspirated engine loses roughly 3.5 percent of its power per 1,000 ft of density altitude and the wing needs a higher true airspeed for the same lift. The FAA’s rule of thumb is to add 15 percent to the computed takeoff distance for every 1,000 ft of density altitude with a fixed-pitch propeller, and 12 percent with a constant-speed propeller.

The FAA’s density altitude pamphlet works a Koch chart example: at a pressure altitude of 6,000 ft and 100°F, takeoff distance to 50 ft increases by 230 percent, so a departure that needs 1,000 ft at sea level needs 3,300 ft, and the rate of climb falls by 76 percent. The pamphlet adds ten percent for high humidity and warns that an aircraft will not perform to book numbers unless the conditions match the book.

The accident record shows what happens when the check is skipped. On 26 July 2012 a Beech B60 Duke departed Sedona, Arizona, where the field sits at 4,830 ft and the density altitude that afternoon was 7,100 ft. The runway is 5,132 ft long. The NTSB calculated that the aircraft should have lifted off 2,805 ft down the runway, and that the distance to accelerate to takeoff speed and then abort safely was about 4,900 ft. It did neither. The Board’s probable cause was the aeroplane’s failure to rotate and the pilot’s failure to reject the takeoff, which resulted in a runway overrun. The pilot and two passengers were killed.

Five weeks earlier, on 30 June 2012, a 1947 Stinson 108-3 had attempted to leave Bruce Meadows, a 5,000 ft turf strip at 6,370 ft in Idaho, with four adults aboard, a computed density altitude around 9,200 ft and a tailwind gusting to 20 knots. A passenger’s camera recorded the whole roll. The aircraft was still on the ground after three quarters of the strip when a gust lifted it; it could not climb, brushed the trees and came down in the forest. Everyone survived. The pilot later told AOPA his density altitude estimate had been about 2,000 ft off and that the departure was not the best decision of his life.

Above: the passenger’s footage of the Bruce Meadows Stinson departure, republished by AVweb. Watch the runway markers go by and the airspeed that never arrives.

How to actually use the rule

Start with the handbook, not the runway. Compute the ground roll for the actual weight, pressure altitude, temperature, wind and surface, apply the density altitude corrections, and add the margin for an engine that is not new. Then find the point on the ground that corresponds to half of that figure. On a paved runway, the FAA design standard for centreline stripes is a 120 ft stripe followed by an 80 ft gap, so each stripe-and-gap pair is 200 ft, though smaller fields sometimes deviate. On grass, pace it out and put down a marker, or note a bush, a fence post or a windsock.

Know the number before you line up. Cavagnaro described asking pilots, as they applied power, what 70 percent of their takeoff speed was, and reported that a sheepish admission of not knowing was the most common answer. For a Cessna 172S with a liftoff speed of 51 knots the target is 36 knots, which is a speed at which the airspeed indicator has barely begun to move. That is the AIM’s first caveat in practice: at 70 percent of liftoff speed, the instrument may not be telling you much.

Treat it as a red light, not a green one. Meeting 70 percent at the midpoint does not certify the takeoff; failing to meet it forbids it. And brief the abort before every departure the way airline crews do, because the decision is far easier to execute if it has already been made on the ground. Brown’s advice, from a career of fighter departures, applies just as well to a Cub on a gravel bar.

The 70/50 rule is not wrong. It is incomplete, and it is only as good as the takeoff distance it is measured against. Used as the FAA Safety Team describes, tied to a computed distance, with an obstacle variant when trees are involved and with the number memorised before the roll, it catches the departure that was never going to work while there is still room to stop. Used as a vague reassurance about the runway as a whole, it is exactly what its critics say it is.

Above: the AOPA Air Safety Institute’s short primer on choosing an abort point before takeoff.

Sources: FAA Aeronautical Information Manual, section 7-6-8; FAA Safety Team (FAASTeam), Safety Enhancement Topic 16-04; FAA pamphlet P-8740-2, Density Altitude; AOPA Pilot (Catherine Cavagnaro, 2019 and 2023; Larry Brown, 2012; James Rush Manley, 2020; Mike Collins, 2016; Barry Schiff, 2000); Boldmethod; Air Facts Journal; mountainflying.com (Sparky Imeson); NTSB report WPR12FA326; NTSB report WPR12LA283 and FAA Lessons Learned, N773C.