Aims. This study examined the effects of two primary care interventions (a physician intervention and a clinic-based psychoeducational group) on drinking patterns, psychosocial problems and blood test results (MCV, GGT, SGOT and SGPT). Design. Subjects were randomized into one of four treatment groups: physician intervention, psychoeducation, both interventions, or no intervention. Follow-up data were collected at 12 and 18 months. Setting. Subjects were recruited from a family practice outpatient clinic managed by a public hospital. Participants. Included 175 Mexican-American female and male primary care patients who screened positive for alcohol abuse or dependence. These patients were not seeking help for alcohol problems. Interventions. Included a brief physician intervention and a 6-week patient psychoeducational group. Measurements. The Diagnostic Interview Schedule assessed subjects for alcohol abuse; the Addiction Severity Index measured alcohol-related problems, including psychosocial issues. Findings. All four treatment groups demonstrated significant improvement over time, with few differences between intervention and control groups. Conclusions. Assessment can be confounded with brief interventions; future investigators should use non-assessed control groups.