Five. Out of 311 routes we track fare history for, five are actually cheapest to fly in July. The other 306 all bottom out somewhere else on the calendar, and for nearly half of them, that somewhere is September or October.
That is the headline sitting inside Flymundus's own fare data. Every route on the site accumulates months of price observations as we log them, and once a route has enough history to say something real, we can ask a simple question: for each route, which month actually produced the lowest fare? The answer settles a debate that gets recycled every summer without anyone checking it against real numbers.
What the data shows
We looked at 311 routes with at least 8 fare observations spread across 6 or more months, the threshold below which a "cheapest month" claim is really just noise from a handful of data points. Here is how those 311 routes broke down by the month their lowest fare actually landed:
| Month | Routes cheapest that month | Share |
|---|---|---|
| September | 80 | 25.7% |
| October | 75 | 24.1% |
| November | 46 | 14.8% |
| August | 44 | 14.1% |
| December | 36 | 11.6% |
| January | 12 | 3.9% |
| July | 5 | 1.6% |
| February | 5 | 1.6% |
| March | 5 | 1.6% |
| April | 3 | 1.0% |
Add September and October together and you get 155 routes, essentially half the set. No other single month, and no other two-month combination, comes close. May and June did not produce a single cheapest fare across all 311 routes. Neither did July, really, aside from the five outliers above.
The gap between the expensive months and the cheap ones is not small, either. Across the 282 routes with usable July and September prices, the median route was 47.7% more expensive in July than in September. Against October, the median premium was 42.7%. Even the smaller shift, August into September, still came out to a 7.4% drop, which on a $400 fare is real money for one month of patience.
Why September and October, specifically
The pattern is not random. School calendars across North America and most of Europe reset in late August or early September, which knocks out the single biggest driver of summer demand almost overnight. Airlines that were running full flights in July suddenly have seats to fill, and fares move fast because pricing systems reprice constantly against actual bookings, not against a fixed seasonal calendar.
October gets a second push from something less obvious: airlines build winter schedules assuming demand keeps falling after the September reset, so capacity on many routes does not tighten again until the November and December holiday run-up begins. That leaves a five to six week window where supply is still generous and demand has not recovered, which is exactly the gap this data is picking up.
Not every route follows the pattern. 44 of the 311 bottom out in August instead, mostly routes where late-summer demand collapses earlier than the September-driven majority, and a handful of long-haul routes to the Gulf and East Asia are cheapest well outside the September-October window entirely. The point is not that every route behaves identically. It is that if you are picking a month to fly without checking your specific route first, September or October is the best default by a wide margin.
Three routes, in detail
London to New York priced at $690 in July across our observations and dropped to $321 in September, a 53% swing on one of the most heavily flown transatlantic pairs there is. The carriers setting that floor, Icelandair and Norse Atlantic UK, are both value operators routing via Iceland and Gatwick rather than the legacy hub-to-hub path, which is part of why the September fare sits as low as it does.
London to Los Angeles is the sharpest swing in the full dataset among routes backed by more than one carrier. July fares averaged $988; by October, the cheapest fare we observed had fallen to $372, a 62% drop. This is not a single discount airline holding the floor, either. Both Lufthansa and British Airways showed up in the October pricing, which makes the swing harder to write off as a one-carrier fluke.
Frankfurt to New York tells a similar story out of a different hub: $690 in July, $315 in September, a 54% drop backed by two separate carriers, TAP Air Portugal and Condor, both appearing at the low end of the September pricing.
All three are long-haul routes with genuine demand swings behind the numbers, not thin routes where one unusual fare skews the whole picture.
How we know this
Every figure above comes from Flymundus's own fare tracking, not a third-party index or a press release. We only pulled routes with at least 8 separate price observations spread across 6 or more months of the year, which cuts the full route list roughly in half and removes anything where a single unusual fare could distort the "cheapest month" label. Every number is the cheapest fare we actually observed for that route in that month, not an average, since averages get pulled around by outlier business-class fares that nobody flying economy will ever see.
Where a route's low fares trace back to a single carrier, and several routes in this dataset do, that is worth knowing before generalizing from it: one airline changing its pricing strategy can move that route's numbers in a way a multi-carrier route will not. Where two or more carriers set the floor independently, as with the Los Angeles and Frankfurt routes above, the pattern sits on firmer ground.
This data updates as new fares get logged, so the exact percentages here will drift slightly over time. The direction will not. Check the current numbers for your own route rather than trust a single month's guess, starting with a live search on the route page for wherever you are actually headed.