Experience with a system of selective and hierarchical reporting of susceptibility to antimicrobial agents in the hospital setting is described. In March 1985 and January 1987 the medical records of all patients with susceptibility test results were reviewed; there were a total of 488 susceptibility reports. Antimicrobial therapy was evaluated for appropriateness on the basis of the reported susceptibility test results. Susceptibility reports would not have affected the choice of antimicrobial agents in the majority of cases because patients had already been discharged, infection had not been documented, or appropriate therapy had already been started. In approximately 40% of cases in which susceptibility reports could have influenced prescribing, physicians chose appropriate initial therapy after susceptibility results became available. If only the instances in which susceptibility reports could have influenced prescribing are considered, then therapy was appropriately changed 12.5% of the time in March 1985 and 24.2% of the time in January 1987. Selective reporting of susceptibility to antimicrobial agents should be viewed as an adjunct to, not a substitute for, other interventions to promote appropriate prescribing in cases of infection.