Why Cancellation Rate Matters More Than On-Time Rate When Choosing an Airline

A 3% cancellation rate sounds small. Over 10,000 scheduled flights, that is 300 trips that never departed — and unlike a 45-minute delay, a cancellation gives you nothing to plan around.

On-time rate gets most of the attention when travelers compare airlines, but it measures only the flights that actually operated. Cancellations are excluded from the on-time denominator entirely under the DOT's Marketing Carrier On-Time Performance filing (BTS Table 236). That means a carrier can improve its reported on-time percentage simply by cancelling its most delay-prone flights. The federal record shows this pattern clearly when you look at both figures side by side.

What the federal definitions actually measure

Under 14 CFR §234.2, a flight is "on time" if it arrives at the gate within 14 minutes of the scheduled arrival. Fifteen minutes or more late is a reportable delay. That rule applies only to flights that operated. A cancelled flight does not enter the on-time calculation at all — it is tracked separately as a cancellation.

The BTS collects both figures from reporting carriers (those operating aircraft with 30 or more seats on domestic routes). The cancellation rate is simply the share of scheduled flights that were cancelled before departure. Diversion rate — flights that landed somewhere other than the scheduled destination — is a third figure reported alongside the other two.

Reading on-time rate without cancellation rate is like reading a batting average that excludes strikeouts.

Why a lower on-time rate can signal a safer choice

Consider two carriers operating the same route over the same 12-month period:

• Carrier A: 85% on-time arrival rate, 3.0% cancellation rate
• Carrier B: 82% on-time arrival rate, 0.5% cancellation rate

Carrier A looks better on the headline number. But if you book 100 flights on Carrier A, the historical base rate suggests roughly 3 of those trips will be cancelled outright — meaning you may face rebooking queues, hotel costs, and missed connections with no arrival at all. On Carrier B, roughly 1 in 200 flights is cancelled. The 3-percentage-point on-time gap costs you an average of a few extra minutes on the flights that do operate. A cancellation costs you the day.

This is not a hypothetical construction. The Airline On-Time Rankings tool pulls the DOT's on-time filings and displays cancellation rate and diversion rate alongside on-time rate for every reporting carrier — precisely because the headline figure alone is not enough to assess operational reliability.

How carriers can inflate on-time rates through cancellations

The mechanism is straightforward. When a carrier's operation is stressed — by weather, crew shortages, or maintenance — dispatchers face a choice: delay a flight or cancel it. A delay counts against the on-time rate. A cancellation does not. There is no formal rule prohibiting a carrier from cancelling a flight it expects to run late, and the DOT filing does not adjust on-time rates to account for cancellation frequency.

This does not mean every carrier with a high cancellation rate is gaming the metric deliberately. Weather-exposed hub airports, thin crew reserves, and older fleets all drive genuine cancellations. But the effect on the reported on-time figure is the same regardless of cause: the cancelled flights disappear from the denominator, and the on-time rate rises.

The DOT's delay-cause reporting breaks cancellations into five categories — carrier-caused, weather, National Aviation System, security, and late-arriving aircraft. Carrier-caused cancellations are the most operationally meaningful because they reflect controllable factors: crew availability, maintenance, and scheduling. A carrier with a high share of carrier-caused cancellations is telling you something about its internal reliability that its on-time rate will not.

A worked example: reading the combined record

Using the DOT's Marketing Carrier On-Time Performance data for the 12 months ending March 2026, suppose two carriers each serve the Chicago O'Hare–Miami corridor:

• Carrier A reports an 84% on-time arrival rate and a 2.8% cancellation rate over that period.
• Carrier B reports an 80% on-time arrival rate and a 0.6% cancellation rate.

At 200 scheduled departures per month (a realistic frequency for a major hub pair), Carrier A cancels roughly 5–6 flights per month. Carrier B cancels roughly 1. The traveler on Carrier B accepts a lower probability of arriving within 14 minutes of schedule on any given operated flight, but faces a much lower probability of not arriving at all.

For a leisure traveler with flexible dates, the on-time gap may be the deciding factor. For a traveler with a connecting flight, a client meeting, or a cruise departure, the cancellation rate is the number that determines whether the trip happens.

What to look for when comparing carriers

• Cancellation rate below 1% over a 12-month period is a meaningful threshold. Rates above 2% warrant scrutiny of the delay-cause breakdown.
• Carrier-caused cancellations as a share of total cancellations matter more than weather cancellations, which no carrier controls.
• Diversion rate is a third signal. A high diversion rate relative to peers can indicate routing or operational stress.
• Route-level data often differs from system-wide averages. A carrier with a 1% system cancellation rate may run 4% on a specific thin route.

The DOT's on-time filings report all of these figures. The Airline On-Time Rankings tool surfaces them together so you can compare carriers on the full picture, not just the headline.

Scope and limitations

All figures on this site come from the DOT's Marketing Carrier On-Time Performance filing (BTS Table 236) and cover U.S. domestic operations of reporting carriers only. No international routes, no live tracking, no predictions. These are historical base rates for the periods stated; they do not forecast individual flight outcomes. This guide is informational only and is not professional travel or legal advice.

Last reviewed: July 2026

Frequently asked questions

Does a cancelled flight count against an airline's on-time rate?

No. Under the DOT's reporting framework, cancelled flights are excluded from the on-time rate calculation entirely. The on-time rate reflects only flights that operated, which is why cancellation rate must be read alongside it.

What cancellation rate should I be concerned about?

A cancellation rate above 2% over a 12-month period is worth examining closely, particularly if a large share of those cancellations are carrier-caused rather than weather-driven. Rates below 1% are common among carriers with strong operational reliability records over the same period.

Can an airline legally cancel a flight to avoid a delay being counted?

The DOT does not prohibit cancellations made to avoid delay reporting, and the on-time metric does not adjust for cancellation frequency. Whether any specific cancellation decision reflects that motivation is not something the public filing reveals — but the aggregate pattern across thousands of flights is visible in the data.

Where does the DOT publish this data?

The Bureau of Transportation Statistics publishes the Marketing Carrier On-Time Performance data (Table 236) monthly. It covers all U.S. domestic carriers operating aircraft with 30 or more seats and includes on-time rates, cancellation rates, diversion rates, and delay-cause breakdowns.

Does cancellation rate vary by route, or is it a system-wide figure?

Both. Carriers report system-wide figures, but the underlying data is flight-level, so route-specific cancellation rates can differ significantly from a carrier's overall average. A hub route with heavy traffic may run far below the system average; a thin regional route may run well above it.