To determine if postextrasystolic changes in systolic time intervals can be used to estimate the severity of resting or provocable left ventricular outflow pressure gradient, we studied the cardiac catheterisation records of 42 patients with hypertrophic cardiomyopathy looking for instances of a single premature beat preceded by a control sinus beat and followed by a postpremature sinus beat. There were 75 such instances in 25 patients. In comparison to the control beat, the pre-ejection period in the postpremature beat was shorter by deltaPEP = -20 +/- 11 ms in 73 of 75 instances, and remained unchanged in two. The ejection time in the postpremature beat was invariably longer by deltaET = 37 +/- 20 ms (range: 10 to 85 ms) and the pre-ejection period/ejection time ratio lower than control by delta(PEP/ET) = -0 . 10 +/- 0 . 05 (range: -0 . 01 to -0 . 25). Total electromechanical systole in the postpremature beat was shorter (11/75), the same (10/75), or longer (53/75) than in the control beat, the overall change being deltaEMS = -18 +/- 22 ms. Both deltaPEP and delta(PEP/ET) correlated poorly with the systolic peak left ventricular-aortic pressure gradient in either the control beat (Gc) or the postpremature beat (Gx), and also with the change in gradient (delta G) from the control to the postpremature beat. In contrast, significant linear correlations were found between delta EMS and either Gc, Gx, or delta G; and also between deltaET and either Gc, Gx, or deltaG. Since internal and external measurements of ejection time are known to be almost identical, the regression equation (deltaG = 1 . 65 delgaET -9) relating deltaET and deltaG should be useful for the non-invasive assessment of the magnitude of provocable left ventricular outflow pressure gradient in patients with hypertrophic cardiomyopathy with spontaneous or externally-induced premature beats.