We studied the relationship between QT interval and RR interval when the heart rate was changed by atrial pacing, atropine, isoproterenol, and exercise in nine healthy adult males. The following results were obtained: (1) QT shortening with an increase in heart rate was minimal in atrial pacing and an equation QT = k4 square root RR was obtained from the QT-RR relationship during atrial pacing (r = 0.64, p less than 0.001); (2) shortening of QT with shortening of RR was least in atrial pacing; (3) during recovery from exercise it was largest and almost coincident with Bazett's formula; (4) Bazett's formula, therefore, is clearly applicable in the evaluation of ECGs recorded at rest or after exercise; (5) Bazett's formula can not be applied in the evaluation of an effect of some drug or maneuver on the QT when such a drug or maneuver can affect both the QT and the RR directly. The RR-dependent QT change can be estimated from the equation QT = k4 square root RR and the difference between a measured QT and the estimated RR-corrected QT would be a genuine effect of the drug or maneuver on the QT at that heart rate. We conclude that atropine, isoproterenol and exercise actually shorten the QT interval.