Florida’s Afternoon Rush Packs 24% of the Year’s Crashes into Three Hours. The Cause Isn’t Fatigue, Glare, or Schools.

By April Karaffa, Director of Business Development and Operations, The Moore Law Firm

Summary

• Florida’s three-hour afternoon rush, 3 PM to 6 PM, produced 165,375 crashes in 2025. That is 24.1% of every crash filed, packed into 12.5% of the day’s clock hours (or 17.6% if you don’t count 11 PM through 6 AM as driving hours).

• The same-length morning rush, 7 AM to 10 AM, produced 102,500 crashes (14.9% of the year). Per hour, that is 1.61 times more crash-dense in the afternoon than the morning. 61% more.

• The pattern is rock-solid across every month. All 12 months sit between 22.9% and 25.1%. Range 2.2 percentage points. Summer months, when schools are out, show the same concentration as school-in months. Summer is actually slightly higher. Schools didn’t cause this.

• Behavioral and sensorial hypotheses fail. Distraction, fatigue (lane-departure), sunset glare, and crash-type composition all show afternoon and morning rush behaving nearly identically. The afternoon block has less glare than the morning, not more. Less single-vehicle. Lower fatality rate. The two windows look like the same kind of crash, with nearly twice as much of it in the afternoon.

• What this is, then: the highest-volume injury hour of the day. Per-trip fatality risk is nearly identical to morning (0.22% vs 0.24%) and far below late-night (1.0% to 1.8%). But absolute injury count is the highest of any 3-hour window in the day, around 56,500 injuries in the 3 PM to 6 PM block in 2025 alone.

• Source: the full 2025 FLHSMV/FDOT preliminary release, all 686,674 records.

Why I went looking

I run operations and intake for a Northeast Florida personal injury firm. Part of my job is reading the firm’s caseload like an instrument panel. When you do that long enough, you start noticing rhythms. One pattern jumped out at me last quarter and would not let go: the timestamps on the cases coming through our door cluster heavily in the afternoon. Almost twice as many at 4 PM as at 8 AM.

Before I trusted that observation, I needed to know whether what I was seeing was specific to our practice or a feature of Florida traffic itself. Years of running an overnight hospital lab teaches you to never confuse a local pattern with a systemic one. So I pulled the full 2025 FLHSMV preliminary crash release. 686,674 reports. Every one of them carrying a timestamp accurate to the minute.

What follows is the path I walked, including the hypothesis I started with that turned out to be wrong, and what is left standing once the obvious explanations are gone.

The headline number, in plain terms

Out of 686,674 Florida crashes filed in 2025, 165,375 fell between 3:00 PM and 5:59 PM. That is 24.08% of the year crammed into 12.5% of the available hours (or 17.6% if you don’t count 11 PM through 6 AM as driving hours). The single busiest hour is 3 PM. 56,115 crashes happen in that one 60-minute window. 8.2% of the year’s traffic violence in 60 minutes.

The morning rush, the same-length 7 AM to 10 AM window, runs 102,500 crashes. 14.93% of the year. Per hour, the afternoon produces about 61% more crashes than the morning rush.

61% more is a wide gap.

The 3–6 PM window holds 24% of the year's crashes, a share comparable to morning rush despite covering fewer hours.

So I started looking for what was producing it.

My first instinct, which turned out to be wrong

The most obvious culprit was schools. Schools dismiss between 2 and 3 PM in most Florida districts. That puts parents, buses, and student drivers on the road exactly when the data peaks. If schools were the driver, summer break, when those streams disappear, should produce a noticeable drop in the 3 to 6 PM share.

I split the year. School-in months: January through May, September through December. Summer break: June and July.

Period	3 PM-6 PM share, all crashes	3 PM-6 PM share, weekdays only
School-in months	23.98%	25.19%
Summer months	24.40%	25.51%
Difference	+0.42 pp	+0.32 pp

Summer afternoons are slightly busier than school-in afternoons, not quieter. The school hypothesis does not just fail. The data tilts the wrong direction for it.

I had to delete a paragraph from my own draft over this. Worth it.

Stepping back: does the pattern hold every month?

If I had been wrong about schools, I needed to know whether anything else seasonal was driving this. Hurricane season pulls cars off the road. Snowbird arrival in October-November shifts demographics. December brings holiday travel patterns. If the 24% concentration were being caused by a specific subset of traffic, I would expect month-to-month wobble.

The data shows almost none.

⚠ VISUAL TODO: Insert visual: Sheet 5 of the data pack - Monthly Stress Test (12-month table with conditional-format gradient on 3-6 PM share)

Month	3 PM-6 PM share
January	23.00%
February	24.09%
March	24.29%
April	25.09%
May	24.55%
June	24.19%
July	24.62%
August	24.41%
September	24.11%
October	24.59%
November	22.85%
December	23.14%

Every single month sits between 22.9% and 25.1%. The full-year range is 2.2 percentage points. Whatever drives this concentration is structural to Florida’s weekday rhythm. It does not turn on and off seasonally.

That is the kind of consistency you almost never see cleanly in seasonal datasets. So I went deeper.

Going deeper: are drivers behaving differently in the afternoon?

If the afternoon spike is not schools and is not seasonal, the next obvious place to look is at the driver. Maybe afternoon drivers are tired from their workday and drifting out of lanes. Maybe the post-lunch attention dip elevates distraction. Maybe sunset glare reduces visibility on the homeward commute. Maybe stop-and-go afternoon traffic produces a different kind of crash than the morning’s commute pattern.

I tested four behavioral and sensorial hypotheses against the data, comparing matched 3-hour windows so the comparison would be honest.

⚠ VISUAL TODO: Insert visual: Sheet 4 of the data pack - Falsification Matrix (six hypotheses with verdict-coded results)

Hypothesis: post-lunch attention dip elevates afternoon distraction.

Result: distracted-flagged crashes at 3 to 6 PM run at 15.8%. At 7 to 10 AM, 14.7%. A 1.15 percentage-point gap. Far too small to drive a 61% volume difference.

Hypothesis: end-of-day fatigue causes more lane-departure crashes.

Result: lane departure at 3 to 6 PM runs at 28.2%. At 7 to 10 AM, 29.2%. The afternoon is slightly LOWER than morning, not higher. The real fatigue signature lives at 2 to 3 AM, where lane departure hits 54%. End-of-workday fatigue does not show up in the afternoon data.

Hypothesis: sunset glare causes an afternoon visibility crash spike.

Result: this one was the most surprising. Glare-flagged crashes at 3 to 6 PM run at 0.34%. At 7 to 10 AM, 1.30%. Sunrise glare beats sunset glare by nearly 4 to 1, every month of the year, including winter. The afternoon block has LESS glare than the morning rush, not more.

⚠ VISUAL TODO: Insert visual: Sheet 6 of the data pack - Glare Heatmap (months x hours showing morning-glare dominance)

Hypothesis: rush-hour stop-and-go produces different crash types.

Result: rear-end share at 7 to 10 AM is 34.5%. At 3 to 6 PM, 34.8%. Single-vehicle share is 17.3% versus 16.0%. The crash types are nearly identical between the two windows.

Four hypotheses. Four failures.

⚠ VISUAL TODO: Insert visual: Sheet 3 of the data pack - Window Comparison (head-to-head morning vs afternoon vs late-night across 15 metrics)

The behavioral fingerprint of an afternoon rush crash matches the behavioral fingerprint of a morning rush crash. Same crash types. Similar distraction. Less glare, not more. Less single-vehicle. Modest aggressive-driving uptick of about half a percentage point. None of those small differences explain a 61% volume gap.

If drivers behaved meaningfully differently in the afternoon, you would see it in the data. You don’t.

Where this points next

After ruling out schools, seasonality, fatigue, attention, glare, and crash-type composition, the obvious behavioral and sensorial explanations are exhausted. What is left is more interesting than disappointing.

The most parsimonious remaining explanation is exposure-based, not driver-based. More vehicles on the road in those hours, overlapping in tighter windows, producing the same kinds of crashes the morning rush does. Just 61% more of them. The weekday-versus-weekend difference (afternoon weekday 25.2% vs weekend 21.8%) backs up this read: when the structural weekday-activity engine partially shuts off, the afternoon spike softens but does not disappear, suggesting the cause is partly commute and partly something more general about weekday afternoon traffic density.

Proving that read directly would require vehicle-miles-traveled (VMT) data joined by hour. FDOT publishes VMT separately from the crash record. With that data layered on the 2025 crash file, you could test whether the afternoon’s crashes-per-vehicle-mile rate matches the morning’s (which would confirm pure exposure) or runs higher (which would mean driving conditions in the afternoon really do produce more incidents per mile, even though the behavioral fingerprints look similar).

A few other directions worth pursuing:

• Per-county hour distribution. Do all 67 Florida counties show the same afternoon spike, or is it concentrated in metro counties with denser commute corridors? The answer tells you whether the pattern is universally Floridian or specifically urban.

• Per-corridor analysis. I-95, I-4, the Florida Turnpike, US-1. Each has its own commuter profile. Identifying which corridors dominate the afternoon load would point at infrastructure decisions.

• Year-over-year persistence. The Derived 2023 dataset exists in our archives. If 2023 shows the same 24% in the same hours with the same monthly consistency, the finding becomes structural to Florida traffic, not a 2025 artifact.

• Vehicle-mile crash rate by hour. The headline question this analysis cannot answer. If afternoon crashes-per-VMT match morning’s, the cause is purely exposure. If they are higher, something about the afternoon is genuinely riskier per mile, even if traditional behavioral signals do not pick it up.

What this dataset alone can prove, defensibly: the afternoon’s 1.61x volume gap over the morning is not behavior. It is not schools. It is not glare. It is not crash-type composition. It is exposure or something close to it. Everything else is something a future analysis can refine.

The signal that matters: injury, not death

The supplementary finding that turns the 24% volume number into something practically meaningful, especially for anyone with a Florida crash injury, is what kind of crash this window produces.

⚠ VISUAL TODO: Insert visual: Sheet 2 of the data pack - 24-Hour Master Profile (full hour-by-hour breakdown with fatality and injury color scales)

Window (3 hours each)	Crashes	Fatality rate	Injuries per crash
7 AM-10 AM (morning rush)	102,500	0.238%	0.319
3 PM-6 PM (afternoon rush)	165,375	0.216%	0.342
11 PM-2 AM (peak late-night)	~22,991	~1.4%	~0.42
Florida statewide average	686,674	0.402%	0.348

Per-trip, the afternoon rush is marginally less dangerous than the morning (0.216% vs 0.238%), but the gap is small and both rush hours run below the Florida statewide average. The genuinely dangerous per-trip hours of the day are late-night. A 2 AM Florida driver faces a fatality risk five to nine times higher than the same driver at 4 PM.

In absolute count, however, afternoon kills more people than morning. 357 fatal crashes in the 3 PM-6 PM window, versus 244 in the 7 AM-10 AM window. The reason is volume, not per-trip risk: with 61% more crashes, even a similar per-crash fatality rate produces more deaths.

The clearer story for most Floridians is on the injury side. The afternoon window produces the highest absolute injury count of any 3-hour window of the day. About 56,500 injuries from 3 PM-6 PM crashes in 2025 alone. That is the statistical fingerprint of stop-and-go traffic at moderate speeds. Multi-vehicle pile-ups. Intersection collisions. Parking-lot exits. People hurt their backs and necks. They go home. They wake up three weeks later and realize something is still wrong.

What you do not see at 3 PM is the high-speed single-vehicle off-road crash at 80 mph. Those concentrate in the late-night window, where Florida’s fatality rate climbs four to nine times the daytime average and is heavily alcohol-driven.

56,500 afternoon-rush injuries. Most of those people walked away. Many of them are still living with consequences they did not notice for weeks.

What this analysis can and cannot support

It can support: the 3 PM to 6 PM window contains 24.1% of Florida’s annual crashes, the pattern holds across every month including summer, the afternoon volume is 61% larger than the same-length morning rush, behavioral and sensorial hypotheses fail to explain the gap, the afternoon block has the highest absolute injury count of any 3-hour window of the day, and per-trip the afternoon rush is similar in fatality risk to the morning rush (slightly lower, in fact) but far below late-night.

It cannot support: any specific causal claim about exactly why afternoon traffic is more crash-prone than morning traffic, any per-driver risk multiplication estimate (the rates are population-level, not exposure-level), or any geographic specificity at the corridor or intersection level without the additional analyses I outlined above.

Frequently asked

What time of day are most Florida crashes?

The 3 PM to 6 PM window. It accounts for 24.1% of all 2025 Florida crashes. The single busiest hour is 3 PM, with 56,115 crashes, 8.2% of the year.

Is morning rush worse than afternoon rush in Florida?

Per volume, no. The afternoon block produces 61% more crashes than the same-length morning rush. Per-trip fatality risk is essentially equal (afternoon is marginally safer per-trip, 0.216% vs 0.238%). Both are well below the state average and far below late-night per-trip risk.

Are afternoon-rush crashes more deadly than the average Florida crash?

No. Afternoon-rush fatality rate at 0.22% runs below the Florida statewide average of 0.40%. The lethal hours are late-night, particularly 11 PM through 4 AM, when per-trip fatality rates run four to nine times the daytime average.

Does the pattern shift in summer when schools are out?

No. Summer months show essentially the same 24% afternoon concentration as school-in months. Summer is actually slightly higher, +0.4 pp on overall and +0.3 pp on weekdays. The school hypothesis was tested and rejected.

Why does the afternoon window have so many crashes?

The crash record cannot answer that conclusively. After ruling out behavioral and sensorial factors, the most plausible mechanism involves overlapping commute, errand, and end-of-shift traffic streams converging in a narrower window than the morning. Confirming this would require per-hour vehicle-miles-traveled data joined to the crash record, which FDOT publishes separately. That join is the next analysis worth running.

Does this affect personal injury cases?

The injury volume in this window is high. About 56,500 injuries came from 3 PM-6 PM crashes in Florida in 2025 alone. These cases are predominantly soft-tissue, whiplash, and back-trauma injuries from stop-and-go and intersection collisions. Documenting symptoms at the scene and following through with medical care matter more for these cases than for late-night high-speed crashes, where the medical record tends to do the documentation on its own.

If you’ve experienced an accident on the road in NE Florida, even if it’s from something as common as an afternoon-rush crash, the Moore Law Firm treats every case with high importance. We secure the medical evidence, the scene evidence, and the witness work these cases demand. So don’t be someone who takes a toll all for nothing. When you are ready to talk, we’ll be ready to fight for you.