Six healthy male subjects were given 0.5mg of digoxin on three separate occasions in the form of an elixir and two different commercially available brands of compressed tablets. Blood levels were measured by a radioimmunoassay technique at various time intervals up to 48hr following drug administration. A previously unreported variable in the radioimmunoassay method was found. Varying amounts of digoxin-free serum added to aqueous samples such as urine or hydroalcoholic solutions led to significant changes in the apparent digoxin concentration measured in the sample. The elixir form gave much higher blood levels than either tablet during the first two sampling times, but the two tablets showed nearly identical blood levels and relative bioavailabilities. A previously reported dissolution rate test method showed wide differences between the two tablets and thus failed to correlate with the similar bioavailabilities observed in vivo.