Documented interventions associated with processing prescriptions in a managed care environment were analyzed, and a bench-mark for interventions was proposed. A retrospective analysis of documented interventions by 31 pharmacies contracted by a managed care organization to serve 22,000 Medi-Cal patients was undertaken. An intervention consisted of identifying any problem related to a prescription, taking action, and recording the problem, action, and outcome on a form. Problems were categorized as drug selection issues, clinical issues, errors in prescription writing, and patient education issues. A similar process was followed for describing actions and outcomes. The data were used as an indicator of prescription-related problems in contemporary pharmacies in a managed care environment; data from another study were used as an indicator of "ideal" pharmacist performance in detecting and resolving problems. The estimated intervention rate when pharmacists meeting ideal performance expectations worked in a managed care environment was set as a benchmark. The economic impact of drug selection interventions was determined as well. Data were retrieved and compiled from 595 (93.4%) of 637 interventions performed in 1995, when 93,483 contract-related prescriptions were processed by the 31 pharmacies. Fifty percent of problems related to drug selection issues. The most frequent action was contacting the presciber (56.1%). The overall intervention rate was 0.7% (range, 0-4.1%). It was estimated that 4% of prescriptions would require an intervention; two pharmacies met this level. Interventions resulting in prescription change realized an average decrease in cost of 65.8%. Community pharmacies under contract with a managed care organization intervened on 0.7% of prescriptions, but the rate of intervention across pharmacies was inconsistent, ranging from 0% to 4.1%; a benchmark of 4% would be reasonable.