Male problem drinkers who accepted a treatment recommendation to take disulfiram were compared to those who refused the drug therapy. Subjects were drawn from a clinical population presenting to an outpatient alcoholism assessment service over a 3-year period. Both univariate and multivariate analyses were employed to investigate which client variables were associated with compliance to the disulfiram recommendation. The most important variable that distinguished the two groups was the percentage of other treatment recommendations adhered to by the client with the disulfiram acceptors being more likely to accept the other treatment recommendations. In addition, clients agreeing to take disulfiram were slightly younger and had higher scores on the Michigan Alcoholism Screening Test (MAST). Variables such as socio-economic status, previous treatment history, other drug use and drinking pattern were not associated with agreement to take disulfiram. The implications of these findings are discussed, particularly their relevance to research on the therapeutic effectiveness of disulfiram.