The relationship between the pharmacokinetics and dynamics of digoxin was investigated using a skin blistering technique that allows experimental access to tissue fluid concentrations. Eight healthy volunteers received digoxin, 1.0 mg, and placebo intravenously according to a double-blind crossover design. Drug concentrations were determined during a 72-hour period in serum, urine, and cantharides blister fluid (CBF). Digoxin levels in the hypothetic peripheral compartments were calculated from serum concentrations. Digoxin effects (total electromechanical systole [QS2c], left ventricular ejection time [LVETc], preejection period [PEPc], QTc time, heart rate, and T wave amplitude) were measured simultaneously. Peak levels in the shallow and deep compartments occurred at 12 1/2 to 20 minutes and 3 hours and the maximum concentration in CBF (2.75 +/- 0.48 ng/ml) occurred at 1 hour. Digoxin effects on QS2c, PEPc, and the ratio PEP/LVET were not related to serum concentrations but were closely related to CBF concentrations (r = 0.90). CBF concentrations were then within the range of serum digoxin concentrations usually associated with the treatment of heart failure. Thus, CBF allows experimental access to active drug concentrations after a single intravenous dose.