The randomized response technique is applied on a voluntary basis for estimating the purchase of smuggled liquor in Norway. Of a total sample of 2330 respondents, aged 15 and over, about one half chose the protective option of the RR-technique. The other half responded directly to the supposed sensitive questions. The estimates from the two groups were weighted together to give population estimates for the percentage of buyers, and the quantities bought, of smuggled liquor. The resulting estimates were found to be considerably higher than estimates from a contemporary survey based on direct response only.