Previous research documenting the association between the level of client disturbance and distress in a significant other has generated two competing causal models. Investigators measuring the level of family burden have suggested that it is the discharged patient who causes distress in the family. Investigators of family factors in patient relapse have suggested that it is the family who disturbs the patient. In this study the relative strengths of these two causal models are compared in a sample of outpatients over a six-month period of time. The data indicate that the significant other is causally predominant in the exacerbation of client disturbance.