Long Beach's Flat, Fast Loop Rewards Those Who Respect the Back Half
Long Beach delivers one of Southern California's most accessible marathon experiences: a road loop through a coastal city, net-neutral in elevation, with a 2025 median finish of 4:40:27 across nearly 4,800 finishers. The 2026 edition arrives on October 11 with cooler, cleaner conditions than last year — a setup that could nudge the field toward its better times.
By the numbers
The 2025 field counted 4,762 finishers, with men clocking a median of 4:32:40 and women 4:58:35. Women made up 1,349 of those finishers. The spread between the top tenth and the median tells the story of a broad, participatory field: the top 10% cutoff sat at 3:32:55, the top 25% at 3:57:55, and the 75th percentile at 5:28:07 — a nearly two-hour window from the front quarter to the back quarter of the race.
How the course runs
The GPS profile is flat in every meaningful sense — -3 ft net change across 26.2 miles, with no sustained climb exceeding 68 ft. The one section worth noting is a short climb of 68 ft between miles 17.1 and 18.0 (averaging 1.6% grade), immediately followed by a 72 ft descent from miles 18.0 to 19.4. That exchange arrives late enough to matter if legs are already fading, but neither pitch is long enough to define the race. This is, fundamentally, a flat road course.
What the pace data makes clear is that the course doesn't slow the field — the distance does. The fade curve from 2025 is instructive:
| Checkpoint | Mile | Field Median Pace |
|---|---|---|
| Early miles | 5.5 | 9:01 /mi |
| Half marathon | 13.1 | 9:29 /mi |
| Mile 20 | 20.0 | 11:55 /mi |
| Finish | 26.2 | 11:59 /mi |
The field runs the first 5.5 miles at 9:01/mi, eases to 9:29/mi through the half, then drops sharply to 11:55/mi between miles 13 and 20 — a slowdown of more than two and a half minutes per mile from the opening segment. The final 6.2 miles hold at 11:59/mi, suggesting the field stabilizes rather than collapses in the closing stretch.
Pacing patterns
98% of 2025 finishers ran a positive split, with an average second-half slowdown of 29.6%. Only 35 runners ran negative splits. That figure sounds stark, but it reflects what marathon fields do at every distance: go out feeling good, pay for it after mile 18. The Long Beach numbers are consistent with mass-participation marathons broadly. For runners planning their 2026 effort, the fade curve above is the most useful pacing guide available — the field's first-half pace is the number most likely to be set too ambitiously.
The weather question
Race day 2025 brought 62–72°F conditions with 73% humidity and a start temperature around 72°F — warm enough to compound late-race fatigue. The 2026 forecast calls for 56–66°F, ~61°F at the start, 88% humidity, and winds around 5 mph. The start temperature drops roughly 11 degrees, which historically benefits marathon performance in the opening miles and reduces the thermal load through the back half. The higher humidity is worth monitoring, but the cooler air is the dominant variable — and it favors the field.
What it takes
| Benchmark | 2025 Time |
|---|---|
| Men's winner | 2:26:31 |
| Women's winner | 2:55:13 |
| Top 10% | 3:32:55 |
| Top 25% | 3:57:55 |
| Median | 4:40:27 |
| 75th percentile | 5:28:07 |
The course is Boston-qualifier eligible. Breaking 3:57:55 places a runner in the top quarter of the field; cracking 3:32:55 puts them in the top tenth.
What to watch
Whether cooler conditions move the median. An 11-degree drop in start temperature is a genuine variable — watch whether the 2026 median dips below the 4:40:27 benchmark set in warmer air.
The depth of the positive-split problem. With 98% of the field running positive in 2025, any meaningful shift toward even splits would represent a notable change in how runners approach the back half.
Field size trajectory. At 4,762 finishers, Long Beach draws a substantial regional field. Whether that number grows or contracts in 2026 will reflect the race's standing among fall Southern California options.
AI race preview · generated from official results of prior editions
