Three categories of traffic accidents were analyzed: fatal accidents, single-vehicle accidents, and all traffic accidents. The findings show that the intervention was followed by a significant decrease in the number of traffic accidents. For all traffic accidents, the reduction was approximately 7 percent; for single-vehicle accidents, 11 percent; and for fatal accidents, 10 percent. Although this outcome agrees with previous experience and some other data that suggest a downward trend in the BAC- level, it is possible that the presumed intervention effect is confounded by other factors. The results should therefore be cautiously interpreted. The intervention models included the proxies for mileage and per capita alcohol consumption as control variables. This also made it possible to estimate the fraction of traffic accidents attributable to alcohol. The etiological fraction for fatal accidents was 38 percent, which is markedly higher than that suggested by corresponding clinical findings. Possible reasons for this discrepancy are discussed. 5 tables, 5 figures, and 36 references