The estimation of serum ethanol concentration by measurement of serum osmolality has been re-evaluated. A formula for calculation of serum osmolality was first validated in 193 patients who had not ingested ethanol. The mean difference between measured and calculated osmolality ("osmolal gap") was 1.5 mOsm/Kg +/- 5.3 mOsm/Kg (1 S.D.). In 37 sera from 35 patients who had ingested ethanol, the osmolal gap was highly correlated with measured serum ethanol (r = 0.994). The serum ethanol (mmol/L) was related to osmolal gap (mOsm/Kg) by the formula: Ethanol = 0.83 X osmolal gap. The factor 0.83 indicated that the behavior of ethanol in serum did not conform to that expected from ideal solutions. This equation accurately predicted serum ethanol in 32 additional samples (r = 0.988). We propose that the formula above be used in place of those that have been previously proposed. The previous formulas involved unwarranted assumptions of ideality and had not been experimentally derived from studies of ethanol in serum.