The purpose of this project was to determine the relationship between adverse drug outcomes and knowledge of drugs, quality of a counseling relationship, and perceptions about drugs in an ambulatory clinic population. The presence of these three factors and other indicators of adverse outcomes were identified through a structured interview of 78 patients. Medical records were reviewed 3, 6, and 12 months after the initial interview to determine the presence of adverse drug outcomes. Four outcomes-hospitalizations, unscheduled clinic visits, emergency room visits, and changes in drug regimen-were assessed and categorized according to drug-related problem. Demographics of the group were recorded. Responses to the interview were analyzed to determine differences in counseling relationships between pharmacists and other health professionals. Logistic regression analysis was performed to determine the influence of independent variables on outcome. Relative risks for developing adverse outcomes given the presence of certain independent variables were calculated for significant associations. Significant relationships between predictor variables and adverse drug outcomes were identified for the following independent variables: female gender (RR = 5.2, CI 1.11, 24.31), three or more diseases (RR = 3.3, CI 1.13, 9.75), more drug knowledge and better perceptions (RR = 0.29, CI 0.10, 0.84), perception of no or low drug interference (RR = 0.18, CI 0.044, 0.745), and inadequate data collection by nurses and physicians as perceived by the patient (RR = 3.08, CI 1.09, 8.70). More knowledge and better perceptions about drugs were associated with a reduced risk of therapy changes due to drug-related problems. No association was found between the quality of the pharmacist-patient counseling relationship and adverse drug outcomes.