1. We have studied the absolute bioavailability of three oral formulations of digoxin, 1.0 mg, in 12 young healthy volunteers in a four way randomised cross-over study using an intravenous control. 2. Digoxin tablets (250 micrograms), liquid filled digoxin capsules (100 micrograms) and an experimental enteric-coated capsule (100 micrograms) were evaluated. In vitro dissolution at pH 1 demonstrated extensive hydrolytic breakdown of digoxin from the tablets and capsules but not from the enteric-coated capsules. 3. Serum 'digoxin' concentrations were measured by fluorescence polarization immunoassay (FPI). The systemic availability (+/- s.d.) of the capsules was 70.5 +/- 11.3%, and that of the tablets 71.5 +/- 8.6%. Drug was less available from the enteric-coated capsules (62.1 +/- 10.3%) measured with FPI. These results were reflected in the urinary drug recoveries measured by FPI. 4. By contrast, there were no differences in urinary recovery of unchanged digoxin between any of the oral treatments, when this was measured by h.p.l.c. The cross-reactivity of immunoassays for metabolites of digoxin may produce artefactual results and the optimal pharmaceutical formulation for digoxin remains to be determined.