To compare the demographic characteristics of patients who miss appointments with those who do not and to identify subgroups who would benefit from specific interventions for improving attendance.Retrospective cohort study of an 18-month period.An urban primary care practice.A random sample (N = 477) of patients who were seen at least twice during the study period.Number of missed visits and kept visits, insurer, age, sex, race, ZIP code, and diagnoses.Of the established patients, 48% missed 1 or more visits. Patients in managed care programs, private and Medicaid, were likely to have missed more visits during the study period than those not in managed care programs (P < .001). Medicaid managed care patients had also scheduled more visits. Significantly higher rates of missed appointments were found in patients aged 19 to 35 years (P = .02), blacks (P < .001), patients in Medicaid managed care programs (P < .001), and patients who scheduled more visits (P < .001). After adjusting for age, race, and sex, Medicaid managed care insurance remained a significant (P < .01) predictor of rate of missed appointments.Patients in managed care programs missed more appointments. Patients in Medicaid managed care programs scheduled more appointments and had higher rates of missed appointments than their counterparts in other insurance groups.