OBJECTIVES: To examine patterns of pharmacotherapy and outcomes among patients with chronic heart failure (CHF). METHODS: We identified a national cohort of patients with CHF in the Department of Veterans Affairs (VA) beginning October 1, 1999 (FY00) and tabulated use of cardiovascular agents, 1-year all cause and CHF (DRG 127) hospital use and survival from FY00 through FY02 (September 30, 2002). We determined health care costs from inpatient, ambulatory and outpatient pharmacy services, and calculated the relative share of these costs across VA medical centers. RESULTS: The total number of patients ranged from 261,054 in FY00 to 299,462 in FY02. The average age(sd) ranged from 70.1(10.4) to 71.1(10.4). Therapeutic classes where use increased included: angiotensin-converting enzyme (ACE) or angiotensin II inhibitors (66.7% to 69.3%); beta-blockers (43.2% to 54.0%); statins (42.0% to 51.9%); and spironolactone (9.7% to 12.0%). Classes where use decreased included: digitalis (38.1% to 34.1%) and calcium channel blockers (35.7% to 32.4%). Diuretic use was relatively constant at 73%. All cause and CHF hospitalization decreased from 26.8% to 22.1% and 5.0% to 4.0%, respectively. One-year mortality decreased from 9.6% to 8.7%. Hospital costs per patient at risk for hospitalization (all patients) decreased from $4933 to $4268 (2002). The relative share of costs of care for hospitalization (about half of total costs) thus decreased overall while costs for ambulatory care and pharmacy services increased slightly. CONCLUSIONS: Pharmacotherapy patterns in the VA exhibited evidence-based practice. All cause hospitalization, hospitalization for heart failure and mortality all decreased. Hospitalization costs decreased, resulting in a net decline in the cost of health care delivery for these patients. Findings suggest that provision of good pharmacological care is improving outcomes and shifting the costs of care from inpatient to outpatient pharmacy and ambulatory care.