Use of the usual classification rule for the MAST resulted in 71 percent and 65 percent of the samples classified as 'alcoholic,' far higher than the incidence of alcoholism in studies which used overseas prisoners. Item analyses conducted with these data, and reconsideration of data from previous New Zealand studies using the MAST, indicate that it is a sound detection instrument. The high sensitivity of the MAST by comparison with assessment procedures based on traditional alcoholism criteria, although New Zealand offender groups seem to score higher than comparable overseas offenders and the classification rule may be slightly too liberal. Accordingly, the use of a higher cut-off score seems appropriate when using MAST scores in classifying individuals. A higher cut-off score still resulted in around 50-60 percent of inmates being classified as alcoholics. The likely benefits of rehabilitative efforts within the criminal justice system focusing on alcohol abuse are discussed. (Author abstract)