Furosemide, a potent loop diuretic, is commonly used in the treatment of congestive heart failure (CHF). Unpredictability in the diuretic effect following oral doses has been attributed to variable and incomplete absorption and to variability in the pharmacodynamic response to furosemide. The present study is undertaken to investigate the absorption kinetics and pharmacodynamics of furosemide in patients with CHF during chronic medication. Ten patients with congestive heart failure were maintained on 40 to 160 mg furosemide for a month. The final dose at the end of this period was administered on an empty stomach. Plasma and urine were collected and assayed for furosemide, potassium, chloride, sodium and creatinine. Urine flow was also measured as a function of time. Plasma furosemide concentration-time data were fit to a two-compartment model with either two consecutive, discontinuous first order absorption rate constants or with a single monoexponential input; the former absorption model describing the data better than the latter. Average values of the half-life (205 +/- 28 min) and renal clearance (0.8 +/- 0.09 ml/min/kg) were similar to those reported by previous investigators. Drug excretion-response curves were lower and shifted to the right compared to data reported for normal subjects. Furthermore, a clockwise hysteresis was evident indicating acute within-dose tolerance.