Objective-To investigate abnormalities of skeletal muscle metabolism in patients with congestive heart failure. Setting-A university teaching hospital. Methods-43 patients (22 New York Heart Association (NYHA) grade II, 21 grade III) and 10 controls were studied. A forearm model of muscle metabolism was used, with a cannula inserted retrogradely into an antecubital vein of the dominant forearm. Maximum voluntary contraction (MVC) was measured using handgrip dynamometry. Subjects performed handgrip exercise, 5 s contraction followed by 5 s rest for 5 min at 25%, 50%, and 75% of MVC or until exhaustion. Blood was taken at rest and 0 and 2 min after exercise for measurement of lactate and ammonia. After 30 min the procedure was repeated with fixed workloads of 7 kg, 14 kg, and 21 kg. Results-MVC (kg, mean (SEM)) was lower in patients than in controls (control 42 . 45 (2 . 3); NYHA II 34 . 13 (1 . 3), P = 0 . 003; NYHA III 33 . 13 (1 . 94), P = 0 . 008). Resting lactate (mmol/1) was higher in patients than controls (control 0 . 65 (0 . 06); NYHA II 0 . 84 (0 . 08), P = 0 . 13; NYHA III 1 . 18 (0 . 1), P = 0 . 002). Resting ammonia (mu mol/1) was higher in NYHA III (65 . 7 (6 . 0)) than in NYHA II (48 . 0 (3 . 7), P = 0 . 016); no difference was found between controls (48 . 0 (7 . 1)) and patients. The overall lactate and ammonia response to exercise was greater in NYHA III than in NYHA II and controls (P < 0 . 05). At volitional exhaustion, peak lactate (mmol/1: NYHA III 3 . 31 (0 . 26); NYHA II 2 . 56 (0 . 16); controls 2 . 71 (0 . 22); P = 0 . 022 NYHA III v NYHA II) and ammonia mu mol/1: NYHA III 126 . 4 (8 . 97); NYHA II 92 . 9 (7 . 23); controls 109 (16 . 3); P = 0 . 006 NYHA III v NYHA II) were higher in severe congestive heart failure. Conclusions-Skeletal muscle metabolism is abnormal at rest in congestive heart failure. During exercise, the degree of metabolic abnormality is related to the symptomatic status of the patient.