Fresno's Flat 10K Carries a Hidden Cost: The Return Leg Slows the Field by Nearly a Minute Per Mile
The Two Cities Marathon 10K looks like a speed-friendly proposition on paper — a flat, out-and-back road course in Fresno with calm winds — but the 2025 field told a more nuanced story. The net elevation profile tilts subtly against runners on the way back, and the median finisher paid for it with a pace that degraded from 10:13/mi at mile one to 11:04/mi by the finish line.
By the numbers
The 2025 edition drew 819 finishers, with women making up a clear majority at 477 — roughly 58% of the field. The overall median finish was 1:08:00, with men's median at 1:01:33 and women's at 1:12:42. With only the single 2025 edition provided, field-size trends across years cannot be assessed, but the gender composition suggests a broad recreational base rather than a competitively skewed field.
Where the course bites
The out-and-back layout is deceptively asymmetric. Miles zero through three offer a gentle net downhill — losing 8 feet in the first mile, another 4 feet in the second, and essentially flat through mile three. The return trip is a different story: the course climbs 4 feet from mile four to five and another 6 feet from mile five to six, meaning runners are fighting a modest but consistent grade precisely when fatigue is setting in.
The fade curve reflects this exactly.
| Checkpoint | Mile | Field Median Pace |
|---|---|---|
| Mile 1 | 1.0 | 10:13/mi |
| Finish | 6.2 | 11:04/mi |
That's a 51-second-per-mile deterioration between the opening mile and the closing segment — significant for a race this short. No mid-course checkpoints were recorded, so the precise mile at which the field begins to slip isn't pinpointed, but the elevation data points squarely at miles four through six as the likely culprit.
Pacing reality check
With only two checkpoints in the provided data, a formal positive/negative split breakdown isn't available. What the fade curve does confirm is that the field as a whole goes out faster than it comes back — the opening mile's 10:13 pace is meaningfully quicker than the finishing segment's 11:04. On a course this short, that gap suggests a significant portion of the field is not running even splits.
The weather question
Race day 2025 saw temperatures of 64–74°F with clear skies, light wind around 3 mph, and 53% humidity — warm for a November morning but manageable. The 2026 forecast projects 69–79°F under the same clear, calm, 53%-humidity conditions. The upper end of that range is 5 degrees warmer than 2025, which on a flat course with no shade cover and a second half that already trends uphill could push finishing times slightly higher than last year's medians. The conditions aren't extreme, but runners who went out at an aggressive 10:13 first mile in 2025 may find the calculus less forgiving in 2026.
What it takes
| Benchmark | Time |
|---|---|
| Men's winner (2025) | 0:34:42 |
| Women's winner (2025) | 0:39:38 |
| Top 10% cutoff | 0:52:49 |
| Top 25% cutoff | 0:58:46 |
| Median (overall) | 1:08:00 |
| Median (men) | 1:01:33 |
| Median (women) | 1:12:42 |
| 75th percentile | 1:22:56 |
Breaking 52:49 puts a runner in the top tenth of the field. The gap between the top 25% cutoff (58:46) and the overall median (1:08:00) — more than nine minutes — illustrates how wide the performance spread is across this field.
What to watch
Can the field hold its opening pace? The 51-second fade from mile one to the final segment is the defining characteristic of this race. Whether 2026's slightly warmer forecast widens that gap will be the clearest story in the results data.
The women's majority. With women comprising 58% of 2025 finishers, the women's competitive depth — and whether the women's winning time of 39:38 holds or falls — is as compelling a storyline as the men's race.
Temperature creep. A high of 79°F would represent the warm end of what this course has seen. How the median finish time responds to that extra heat, on a return leg that already climbs, will be worth tracking once results post.
AI race preview · generated from official results of prior editions
