Using a modified GLC method for the assay of phenacetin, extracts of normal human urine regularly showed a peak with an identical retention time to N-acetyl-p-aminophenol (NAPA). The peak was considerably accentuated after ingestion of ethanol. Reanalysis of the derivatized extracts after spiking them with small amounts of NAPA yielded a larger peak having a mass spectrum clearly that of a mixture of NAPA and p-HPAA TMS [p-hydroxyphenylacetic acid-trimethylsilyl] derivatives. When accurate measurement of NAPA in urine is required at low concentrations, as in pharmacokinetic and other studies of the metabolism of phenacetin or NAPA, the errors due to the p-HPAA peak could be serious, particularly if alcohol were taken during the study.