The effect of cardiovascular drugs on the results of the signal-averaged electrocardiogram (SA-ECG) was studied in healthy young subjects. Cardiovascular drugs may confound the analysis of SA-ECG, but the effect of drugs on time-domain analysis and on spectral temporal mapping (STM) in healthy subjects has not been investigated under standardized conditions. Also, the reproducibility of the various spectral area measurements of STM has not been assessed simultaneously. Short-term drug effects on the SA-ECG were studied in 20 healthy young men (age 23.5 +/- 2.5 years) in an observer-blinded, crossover design on separate days. Each subject was randomly assigned to four drugs. The SA-ECG was recorded at baseline and during i.v. administration of stepwise increased doses of adenosine, atropine, isoproterenol, lidocaine, norepinephrine, propranolol, verapamil and placebo. Low amplitude signal duration was significantly and dose-dependently shortened by isoproterenol, whereas no consistent effect on the filtered QRS duration or on the root mean square voltage in the terminal 40 ms of time-domain analysis was noted for any of the other drugs. Only lidocaine significantly and dose-dependently decreased the parameters of STM at QRS offset and +20 ms, even though short-term reproducibility and day-to-day reproducibility of spectral area measurements were low. Interpretation of "normal" results from SA-ECG is not relevantly influenced by the studied drugs at clinical doses. STM may prove a useful technique for detection of the actions of sodium channel-blocking drugs. Reproducibility of STM measurements is substantially lower than that of time-domain parameters.