The purpose of this study was to measure the ability of telemonitoring to reduce hospital days and total costs for Medicare managed care enrollees diagnosed with heart failure. Patients were recruited and randomly assigned for six months to either telemonitoring or standard care. Telemonitoring transmitted vital signs and clinical alerts daily to a central nursing station. Utilization of covered services was analyzed for the six-month telemonitoring period to test for hypothesized reductions in hospital days and changes in utilization of the emergency department (ED), urgent care, and primary care. Negative binomial regressions adjusted for gender, age, co-occurring diabetes, co-occurring chronic obstructive pulmonary disease, and residence neighborhood were used to analyze units of service, and two-part (hurdle) multivariable models were used for expenditures. The main finding was a tendency for lower total number of hospital days for patients assigned to telemonitoring. Results for other covered services were generally consistent with hypothesized direction and magnitude; however, statistical power was reduced because of lower-than-expected recruitment rates into the study. Within a managed-care environment, telemonitoring appears to facilitate better ambulatory management of heart failure patients, including fewer ED visits, which were offset by more frequent primary care and urgent care visits.