To describe the clinical course of patients admitted to a nonintensive care telemetry unit and to determine whether telemetry identifies patients at risk for transfer to the intensive care unit (ICU), 467 patients hospitalized for cardiac monitoring in a nonintensive care telemetry unit were followed until death or discharge. The American College of Cardiology guidelines for telemetry use were applied: 65% of patients were class I (monitoring definitely indicated); 33% class II (probably indicated); and 2% class III (not indicated). In 5 patients (1%), telemetry contributed to the decision for a transfer to the ICU. In 462 patients, telemetry added no significant information. Thirty-eight patients (8.1%) were transferred to an ICU: 22 because of cardiac deterioration and 16 because of noncardiac clinical deterioration. Eighteen percent of patients in class I (95% confidence interval [CI], 14.1 to 22.8), 12% in class II (95% CI, 6.7 to 17), and none in class III (95% CI, 0 to 26) were transferred to the ICU (p = 0.03). Nine patients died (1.9%), 4 with terminal illness. Three patients died while on telemetry: 1 had metastatic lung cancer and 2 died suddenly of cardiac causes during initial evaluation on the ward. Telemetry identified the terminal rhythm in the 3 patients. Patients admitted to a non-ICU monitored ward with ischemic syndromes, heart failure, and arrhythmia rarely deteriorated. Patients who did deteriorate were recognized clinically without appreciable contribution from the monitoring process. It remains unproven that heart rhythm monitoring in general practice units improves patient care. © 1994.