A meta-analysis has been conducted on the effects on crashes of DUI-checkpoints (DUI, driving under the influence). The results indicate that crashes involving alcohol are reduced by 17% at a minimum and that all crashes, independent of alcohol involvement, are reduced by about 10-15%. In a moderator analysis the effects of a number of factors that may affect the effectiveness of DUI-checkpoints were investigated by means of subgroup analyses and meta-regression. Those moderator variables that were found to be most relevant, are the time period studied, country, and study design. DUI-checkpoints were found to be most effective during the first half year. Australian checkpoints were found to be more effective than checkpoints in other countries. Smaller crash reductions were found in studies that have applied a control group than in other studies. Testing all drivers who are stopped at a checkpoint may improve the effectiveness of DUI-checkpoints. The results do not indicate that DUI-checkpoints have greater effects on more severe crashes or that the use of paid publicity improves the effectiveness. Most likely there are further factors that affect the effectiveness of DUI-checkpoints that could not be investigated in the present analysis.