A multifaceted intervention was hypothesized to increase postdischarge ambulatory contacts and to reduce nonelective readmissions. Patients (N = 1,001) discharged from the general medicine service were stratified by risk for nonelective readmission and randomized to the control or intervention group. Intervention patients received phone calls from nurses, mailings of appointment reminders and lists of early warning signs, and prompt rescheduling of visit failures. Patients were followed for 6 months, and the results were computed in units per patient per month. The intervention group had 10.4% more total office contacts (0.53 vs 0.48, P = 0.005) than the control group. Although the intervention group had 7.6% fewer nonelective readmission days (0.85 vs 0.92), this was not statistically significant (P = 0.5). Patients in the intervention group at high risk (N = 181) had 28.1% more office visits (0.73 vs 0.57, P less than 0.01) and 31.9% fewer nonelective readmission days (1.13 vs 1.66), but this was also not statistically significant (P = 0.06). Thus, the intervention significantly increased post-discharge contacts, primarily in high-risk patients, but failed to reduce the incidence of nonelective readmission days significantly.