In a randomized clinical trial of Inpatient Family Intervention (IFI) for 169 inpatients with schizophrenia, affective disorder, and a residual group of other diagnoses, results suggested significant effects favoring IFI for patients and their families. The treatment effects were limited to females and to two diagnostic groups: chronic schizophrenia patients and the bipolar subgroup of affective disorders. Using analysis of variance and correlational analyses, we examined three variables that mediate between family treatment and outcome: 1) posthospital medication compliance, 2) posthospital psychosocial treatment compliance, and 3) patient rejection by the family. Results showed robust correlations between these variables and outcomes for all diagnostic subgroups. The pattern of associations between the mediating variables and outcome was more prominent for patients treated with IFI. These results were seen most clearly in the total sample, and for the subgroups of “all females” and “poor prehospital functioning females with schizophrenia.” Finally, family achievement of the goals of the family intervention were positively associated with better patient outcome, while increased family burden at discharge was associated with shorter length of time posthospital to rehospital