BACKGROUND The effects of anaesthetics on left ventricular (LV) diastolic function in patients with pre-existing diastolic dysfunction are not well known. We hypothesized that propofol but not sevoflurane will worsen the pre-existing LV diastolic dysfunction. METHODS Of 24 randomized patients, 23 fulfilled the predefined echocardiographic criterion for diastolic dysfunction. They received general anaesthesia with sevoflurane 1 MAC (n=12) or propofol 4 mug ml(-1) (n=11). Echocardiographic examinations were performed at baseline and in anaesthetized patients under spontaneous breathing and under positive pressure ventilation. Analysis focused on peak early diastolic velocity of the mitral annulus (E(a)). RESULTS During spontaneous breathing, E(a) was higher in the sevoflurane than in the propofol group [mean (95% CI) 7.0 (5.9-8.1) vs 5.5 (4.7-6.3) cm s(-1); P<0.05], reflecting an increase of E(a) from baseline only in the sevoflurane group (P<0.01). Haemodynamic findings were similar in both groups, but the end-tidal carbon dioxide content was more elevated in the propofol group (P<0.01). During positive pressure ventilation, E(a) was similarly low in the sevoflurane and propofol groups [5.3 (4.2-6.3) and 4.4 (3.6-5.2) cm s(-1), respectively]. CONCLUSIONS During spontaneous breathing, early diastolic function improved in the sevoflurane but not in the propofol group. However, during positive pressure ventilation and balanced anaesthesia, there was no evidence of different effects caused by the two anaesthetics.