Serum digoxin concentration (SDC) was compared with clinical and ancillary predictors as a guide to adjustment of digoxin dose and as a test for digitalis toxicity in a total of 76 hospitalized patients during a period of 9 months. The mean SDC (3.6 +/- 2.5 nmoles/liter) associated with unexpected discontinuation of therapy was significantly higher (p less than 0.001) than that (1.1 +/- 0.6 nmoles/liter) associated with unaltered digoxin dose, while the mean SDC (0.6 +/- 0.4 nmole/liter) associated with unexpected dose increase was significantly lower (p less than 0.05). There was no significant association between other pharmacokinetic or pharmacodynamic predictors and therapeutic intention. There was a 13% incidence of confirmed digitalis intoxication. The mean SDC (3.6 +/- 1.9 nmoles/liter) of patients presenting and confirmed as digitalis toxic was significantly higher (p less than 0.001) than that (1.4 +/- 0.6 nmoles/liter) involving a situation in which digitalis toxicity could not initially be excluded by other means. The predictive value of an SDC greater than or equal to 2.6 nmoles/liter for toxicity was 80%, and its efficiency for diagnosing both toxicity and nontoxicity was 95%. The SDC was thus shown to be a valid test of digitalis toxicity and to provide extraordinary information enabling the clinician to modulate digoxin therapy precisely.