The 9-minute self-powered treadmill test has been employed to evaluate submaximal exercise capacity in heart failure patients, but its relation to maximal exercise capacity and to indexes of skeletal muscle function has not been well defined. Two protocols were utilized. The first evaluated the relation of the peak oxygen uptake (VO2) achieved on the self-powered treadmill to that during a symptom-limited treadmill protocol, and examined the reproducibility of this test. Thirteen patients (aged 62 +/- 2 years, in New York Heart Association class I to III [2.3 +/- 0.1], ejection fraction 23 +/- 2% [means +/- SEM]) and 10 age-matched sedentary controls were studied. The second protocol, which involved 18 patients (aged 65 +/- 2 years, in New York Heart Association class I to IV [2.4 +/- 0.1], ejection fraction 23 +/- 2%) and 10 age-matched controls evaluated the relation of performance on the self-powered treadmill to maximal systemic exercise capacity on a cycle ergometer and to indexes of skeletal muscle function. In the first protocol, the test was found to be highly reproducible. The proportion of self-powered treadmill to maximal treadmill peak VO2 did not differ significantly between patients and controls (95 +/- 5% vs 87 +/- 6%). In the second protocol, patients achieved a lower peak VO2 (15.6 +/- 1.1 vs 25.6 +/- 0.9 ml/kg/min, p < 0.001), walked a shorter distance on the self-powered treadmill (367 +/- 32 vs 667 +/- 28 m, p < 0.001), and exhibited less knee extensor work capacity (1,075 +/- 116 vs 1,390 +/- 110 ft-lbs, p < 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)