Somewhere between a quarter and a third of DUI arrests in the United States involve a repeat offender. The percentage of DUI arrests that are repeat offenders depends heavily on how a state defines a prior offense, how far back its records reach, and whether the count includes arrests, convictions, or both. The most cited federal estimate puts the share at roughly one-third; newer analysis puts it closer to 25 percent.
What the National Data Shows
The National Highway Traffic Safety Administration’s widely referenced estimate placed repeat offenders at approximately one-third of the 1.5 million annual drunk driving arrests in the country.1NHTSA. Guide to Sentencing DWI Offenders That figure dates to 1995 and has served as the baseline for policy discussion ever since. Later NHTSA research suggested the proportion had declined to roughly 25 percent, though that estimate carries caveats about tracking methods and timeframes.
The gap between those two numbers is real but partially an artifact of measurement. The one-third figure counted anyone with any prior DUI-related contact. The lower estimate applied narrower criteria. Both are snapshots. The true national rate almost certainly falls somewhere in between depending on the state, the time period studied, and whether the data counts arrests, convictions, or both. Individual state reports have found repeat-offender shares ranging from the low twenties to about 30 percent in any given year.
Fatality data tells a related but narrower story. In 2023, 12,429 people died in alcohol-impaired driving crashes across the country.2NHTSA. Drunk Driving Statistics and Resources Research has found that roughly 8 percent of intoxicated drivers involved in fatal crashes had at least one prior DWI conviction within just the previous three years. That three-year window almost certainly undercounts the true number, because many convictions fall outside such a narrow lookback.
Why the Numbers Vary So Much
If you compare repeat-offender statistics across states or studies, the figures will rarely match. That isn’t because the data is wrong. It’s because the definitions aren’t the same.
Look-Back Periods
Every state sets a look-back period that determines how far into the past a prior DUI counts for penalty enhancement. These windows range from five years in some states to a lifetime in others. Illinois and Texas, for example, use lifetime look-back periods, meaning a DUI from 20 years ago still counts as a prior offense. Arizona uses seven years. California, Georgia, Ohio, and Virginia use ten. A state with a lifetime look-back will naturally report a higher share of repeat offenders than a state that only counts priors from the last five years, even if actual driver behavior is identical.
Arrests Versus Convictions
Whether a study counts arrests or convictions changes the numbers significantly. An arrest means a law enforcement officer had probable cause to take someone into custody for impaired driving. A conviction means a court found the person guilty or the person pleaded guilty. Many DUI arrests end in plea bargains to lesser charges like reckless driving, which may not show up as a prior DUI in some databases. Studies that track arrests capture a broader set of repeat behavior than those limited to convictions.
Cross-State Tracking Gaps
Someone arrested for DUI in one state may have prior convictions in another state that don’t appear in local databases. The National Driver Register exists to help states share information, but data sharing remains inconsistent. The national repeat-offender rate is almost certainly underreported for this reason, because some prior offenses simply aren’t visible to the arresting jurisdiction.
Why the Share Matters
A relatively small group of repeat offenders accounts for a disproportionate share of impaired driving arrests and an outsized share of the danger on the road. Alcohol-impaired crashes killed more than 12,000 people in 2023, and drivers with prior DUI records were involved in a meaningful fraction of those deaths.2NHTSA. Drunk Driving Statistics and Resources That concentration of risk is why federal and state policy has focused so heavily on this subset of drivers rather than on first-time offenders alone.
One point worth being clear on: the 25-to-33 percent figure describes the share of DUI arrests involving a repeat offender, not the share of individual drivers who re-offend after their first arrest. Those are different measurements, and the file does not report a national lifetime recidivism rate for people arrested once.
Why the Number Has Trended Down
The repeat-offender share appears to have declined from the one-third level measured in the mid-1990s. Two interventions explain most of the movement.
Ignition interlock devices, which require a driver to pass a breath test before the vehicle starts, have become the most widely adopted tool for preventing repeat drunk driving. Thirty-one states and the District of Columbia require them for all DUI offenders. An additional eight states require them for repeat offenders and those with high blood-alcohol levels, and five more mandate interlocks exclusively for repeat offenders. According to the CDC, ignition interlocks reduce DUI recidivism by about 67 percent compared to license suspension alone. The protection fades once the device is removed, which is why many states now require installation periods of two or more years.
Specialized DUI courts have shown the strongest results of any intervention studied. The most conservative studies show these courts reduce recidivism by about 12 percent, while the best-performing programs cut it by as much as 60 percent. One study found that DUI court participants were 19 times less likely to be re-arrested for drunk driving within two years compared to offenders processed through traditional courts. The common ingredient is sustained accountability: regular judicial check-ins, swift consequences for violations, and treatment long enough to change behavior.
License suspension on its own is a surprisingly weak deterrent. Many repeat offenders drive on suspended licenses anyway. Jail time without treatment also shows limited effect on preventing future offenses, because the underlying issue for many repeat offenders is alcohol use disorder, and incarceration alone doesn’t address it.
The Takeaway
Whether the true national share sits at 25 percent or closer to one-third, the core finding is consistent: roughly one in four to one in three DUI arrests involves someone who has been through the system before. The trend has moved in the right direction over the past three decades, driven by interlocks, DUI courts, and tougher state penalty structures. But the gap between the current rate and what sustained intervention could achieve remains wide.