Responses came from 391 of the 400 residents who received the mailed questionnaire containing vignettes of different situations involving drunk driving. Two hundred and five of the surveys were usable, for a response rate of 52.4 percent. Results revealed that harm, causing a fatal injury, and one measure of culpability (the drunk driver's prior arrest record) significantly increased punitiveness ratings. However, other indicators of offenders' culpability do not affect the sanctions imposed by the study participants. Findings suggested two policy implications. First, it appears that harm, not recklessness, largely determines how severely citizens believe a drunk driver should be punished. Second, findings suggest that current legal policies that consider first-time drunk drivers as misdemeanants and repeat offenders as felons are largely consistent with the public will. Tables, footnotes, and 56 references