An open look at one year on Arizona's roadsData Year 2024 · Free to read, cite & share
Phoenix · Maricopa County · Arizona
Phoenix's Deadliest Intersections
What a year on Arizona's roads actually looks like, by the numbers, and what the record says about when and where a crash turns fatal.
1,228
people killed on Arizona roads in 2024.
THAT'S 3.36 EVERY DAY. ONE LIFE EVERY 7 HOURS, 9 MINUTES.
Who reads the data
We read the crash record the way we read a case: the road, the numbers, and the pattern underneath them.
AZ Law Now reads Arizona's crash record. We pulled the state and federal data and computed the lethality ourselves: which behaviors, which hours, and which corners actually turn a crash into a death.
It names no one and sells nothing. It just shows the work.
Why this matters now
On June 9, 2026, a suspected impaired driver ran a red light at more than twice the posted limit in Phoenix and caused a fatal multi-vehicle crash. Three of the deadliest things a driver can do, in one moment. The data shows it's a pattern, not an accident of one bad night.
01
Not all crashes kill equally
Our analysis
Speeding causes the most crashes by far. But per crash, it isn't the deadliest act.
We took ADOT's driver-violation table and computed a kill rate for each behavior: the share of drivers with that violation who ended up in a fatal crash. Running a red light is 2.32 times more lethal per crash than going too fast for conditions.
Kill rate by driver behavior ComputedDrivers in fatal crashes ÷ all drivers with the violation
BehaviorLethality per crashKill rateOdds
Crossed the median150 drivers · 15 in fatal crashes
10.00%
1 in 10
Exceeded the lawful speed733 drivers · 71 in fatal crashes
9.69%
1 in 10
Drove wrong-way136 drivers · 13 in fatal crashes
9.56%
1 in 10
Drove left of center1,362 drivers · 88 in fatal crashes
6.46%
1 in 15
Ran a stop sign1,418 drivers · 25 in fatal crashes
1.76%
1 in 57
Ran a red light The intersection behavior5,891 drivers · 64 in fatal crashes
1.09%
1 in 92
Failed to yield23,078 drivers · 139 in fatal crashes
0.60%
1 in 166
Went too fast for conditions32,466 drivers · 152 in fatal crashes
0.47%
1 in 214
Hard speeding is nearly 10x more lethal per crash than "too fast for conditions."
Computed from ADOT 2024 Crash Facts, Table 5-11 (Driver Violations). Kill rate = drivers in a fatal crash ÷ all drivers cited for that violation. Presence in a fatal crash does not establish fault.
The common read is that speed is the deadliest behavior because it causes the most crashes. It isn't.
Speed is the volume leader. Per crash, running a red light kills 2.32x more often than driving too fast for conditions.
The danger at an intersection isn't how many crashes happen there. It's the specific act.
02
The midnight multiplier
Our analysis
The hour of a crash predicts its lethality more than almost anything else. We computed a kill rate for all 24 hours of the day.
The hours with the most crashes and the hours with the most deadly crashes are not the same window. They barely overlap.
Crash volume peaks at rush hour. Lethality peaks after midnight.
Crash volumeKill rate
A crash at midnight has a 3.51% chance of being fatal. The same crash at 4 PM: 0.45%. That's a 7.8x multiplier.
Rush hour fills the roads, but the road empties out and turns lethal after dark: the kill-rate line climbs as the volume line falls. Sunday is the deadliest day (1.79x the Thursday rate), and a weekend crash is 1.74x more lethal than a weekday one.
Computed from ADOT 2024 Crash Facts, Table 3-7 (Crashes by Hour and Day of Week). Kill rate = fatal crashes ÷ all crashes in that hour. Day and weekend rates computed the same way.
03
The intersection signature
Our analysis
There's a crash type that belongs to the intersection: the left turn and the angle collision. We computed the kill rate for every collision manner. A left-turn crash is 3.2 times more likely to be fatal than a rear-end.
Left-turn collision0.75% kill rate · 1 in 133
3.2x the rear-end rate
127 of 16,866 left-turn crashes were fatal. The turn across traffic is the signature intersection crash, and it kills.
Angle (broadside) collision0.53% kill rate · 1 in 189
2.2x the rear-end rate
96 of 18,182 angle crashes were fatal. Together with left turns, angle and left-turn crashes are 1 in 3 multi-vehicle crashes but 28% of the fatal ones.
Rear-end collision0.24% kill rate · 1 in 425
the baseline
91 of 38,654 rear-end crashes were fatal. The most common multi-vehicle crash is also the most survivable.
Computed from ADOT 2024 Crash Facts, Table 3-1 (Manner of Collision in Multi-Unit Crashes). Bars scale to kill rate, left-turn = 100%.
Put the three findings together: a left turn, run on a red light, after midnight. That's the deadliest single moment on a Phoenix road. June 9 stacked all of it.
04
Where: the deadliest intersections
6 in 10 traffic injuries and about 4 in 10 traffic deaths in metro Phoenix happen at an intersection.
MAG, 2018 to 2022 · Maricopa Association of Governments intersection risk analysis (NSM-I)
Highest-risk metro-Phoenix intersectionsSOURCE: MAG · DATA YEAR 2018 TO 2022 · SEVERITY-WEIGHTED
167th Ave & McDowell RdPhoenix
683rd Ave & Indian School RdPhoenix
251st Ave & Camelback RdGlendale
7Cave Creek Rd & Sweetwater AvePhoenix
319th Ave & Peoria AvePhoenix
851st Ave & Thomas RdPhoenix
467th Ave & Thomas RdPhoenix
927th Ave & Camelback RdPhoenix
567th Ave & Indian School RdPhoenix
1099th Ave & Lower Buckeye RdPhoenix
9 of the 10 sit on the west side.MAG intersection risk analysis (NSM-I), 2018 to 2022
Our fresh corroboration Computed
We pulled the federal fatal-crash records, and the west-side pattern holds.
The named ranking above is MAG's, and it covers 2018 to 2022. To check whether it still describes the road today, we downloaded NHTSA's federal fatal-crash files for 2022 and 2023, filtered to Maricopa County, and grouped 1,276 fatal crash events by intersection.
The same arterials dominate both datasets. 19th Avenue, 43rd Avenue, 51st Avenue, and 67th Avenue run through both rankings.
Two of MAG's exact corners reappear among the deadliest in the fresh federal data: 67th Ave & McDowell and Cave Creek & Sweetwater. Two independent sources, two different methods, the same west-side map.
The 43rd Avenue corridor is the new signal. In the fresh federal data, 43rd Ave is the single most lethal arterial: fatal crashes at Thunderbird, Thomas, and Northern, totaling 10 deaths across 5 crash events in 2022 and 2023. These spots repeat across both years, which reads as structural, not chance.
43rd Ave & Thunderbird Rdfatal '22 + '23
43rd Ave & Thomas Rdfatal '22 + '23
35th Ave & Thunderbird Rdfatal '22 + '23
19th Ave & Northern Avefatal '22 + '23
103rd Ave & Olive Avefatal '22 + '23
67th Ave & McDowell Rdin both rankings
Computed from NHTSA FARS 2022 to 2023, Maricopa County, fatal crashes only. Intersections grouped on primary and cross street with a geographic consistency check. Single-name geo-suspect entries excluded.
An honest methodology note
Our corroboration counts fatal crashes only, because that's the federal data anyone can download. A complete current, all-severity ranking, the kind that captures injury and fender-bender crashes too, needs ADOT's CRIS incident-level data, which the state does not publish openly. We've filed for it through a public-records request.
Until it lands, we say plainly what we have: a fresh fatal-crash corroboration that confirms MAG's west-side pattern, not a replacement for it. The June 9 crash is a behavior example only. We make no claim that its location ranks.
05
Maricopa crashes more, dies less
Our analysis
Phoenix isn't just dangerous. It's a crash factory where most crashes don't kill. The danger is the subset that does.
Crashes vs deaths
73%
of Arizona's crashes happen in Maricopa County, but only 48% of the deaths. Outside Maricopa, a crash is 2.65x more likely to kill. Rural interstates and mountain roads turn crashes fatal far more often than dense, lower-speed city streets.
Computed from ADOT 2024 Crash Facts, Table 2-2.
The daily bill
$32.4M
leaves Maricopa County every single day to crashes. The county's crashes cost $11.8 billion in 2024, 56.6% of Arizona's entire crash bill, about $2,600 per resident per year.
Computed from ADOT 2024 Crash Facts, Table 1-4 (Estimated Economic Loss by County).
06
The statewide context
Killed · 2024
1,228
Down 6.12% from 2023, the second straight drop.
ADOT 2024 Crash Facts
Crashes
121,107
Reported across Arizona in one year.
ADOT 2024 Crash Facts
Injured
54,426
149 every day. Roughly one every 10 minutes.
ADOT 2024 Crash Facts
Alcohol deaths
347
28.26% of all road deaths in 2024.
ADOT 2024 Crash Facts
What this means
Every number points the same way. The deadliest crash isn't the busy one.
It's the left turn run on a red light, after midnight, on a west-side arterial. Knowing where the danger concentrates, and when, protects families and the public.
Every figure traces to a named government dataset: ADOT's 2024 Crash Facts for the statewide record, NHTSA's FARS federal fatal-crash files for the fresh intersection corroboration, and the MAG intersection study for the named ranking. We read the source, not the news write-up.
We computed the kill rates
The lethality numbers are ours. ADOT publishes crash counts.
We divided fatal crashes by total crashes per behavior, per hour, per collision type, and per county to get a kill rate nobody else publishes from this data. Every calculation uses only the published counts.
Verified, then flagged
Computations were checked in code against ADOT's own totals. Where a number was thin or a name looked ambiguous in the federal file, we flagged it and left it out.
We did not design a finding. We struck it from the data.
The one limitation we want stated plainly
Our intersection corroboration covers fatal crashes only, because FARS is the federal fatal-crash file. It is the right source for a "deadliest" claim, but it is not an all-severity "most crash-prone" ranking.
That requires ADOT's CRIS incident-level export, which the state does not publish openly and which we've requested through public records. MAG's ranking is severity-weighted but covers 2018 to 2022.
We present these as two complementary sources that agree, and we disclose each one's date and scope rather than blend them into a single false ranking.
Provenance ledger · every headline number
Measured read directly from a published table.
Computed our calculation from published counts.
Verified confirmed against the named source.