RESUMO
<p><b>OBJECTIVE</b>To estimate the number of drug users in Hongjiang, Hunan and to develop strategy for drug reduction in the future.</p><p><b>METHODS</b>Two capture-recapture methods were used to estimate the numbers of drug addicts. Random stratified sampling survey was used to verify the optimum allocation. The first capture-recapture method (CR1) referred to the number from optimum allocation random stratified sampling survey conducted in the communities and the number from local Public Security Bureau list being the second capture. The second capture-recapture method (CR2) referred to the collection of records in the detoxification unit with an interval of 4 months. The estimated number was calculated under Seber's adjustment formula. Face to face interview was carried out during the optimum allocation random stratified sampling survey process.</p><p><b>RESULTS</b>Of 1388 interviewed in the communities, 24 (1.73%) were identified as drug addicts under the optimum allocation random stratified sampling survey. When the figure 1.73% was applied to the total population (72,709) in Hongjiang, the result yielded an estimation of 1258 drug addicts. The estimated numbers of CR1 and CR2 were 904 and 1069 respectively. However, the number was 1.3 to 1.6 fold higher than the reported number (687) by local Public Security Bureau.</p><p><b>CONCLUSION</b>The capture-recapture method seemed a better method in estimating the number of drug addicts.</p>