Men and women entering drug treatment programs are known to differ in demographic characteristics and psychosocial behavioral patterns. To be effective, any such program that caters to individuals from both sexes should identify and address these gender-based variations. Studies have also reported clinical differences in the effect of drugs on men and women addicts. Here, we examine whether the treatment is equally effective on men and women, when several demographic covariates are controlled. We construct a "problem severity index" to categorize individuals based on how acute their problems were at the start of the program. We also examine how the choice of treatment modality affects treatment success. Cumulative logit models were used in our analysis. The choice of treatment modality is sex specific, although sex did not appear to be a significant factor for treatment effectiveness when we controlled for other explanatory variables.