Nisoldipine was administered at 10(-9) M, a dose lacking negative inotropism, to isolated and perfused rabbit hearts submitted to 60 min ischaemia (1 ml.min-1) followed by 30 min reperfusion. The drug was delivered either 30 min before ischaemia, at the onset and after 30 min of ischaemia and during reperfusion only. Cardiac protection was evaluated in terms of recovery of left ventricular pressure during reperfusion, release of creatine phosphokinase (CPK), mitochondrial function, tissue content of adenosine triphosphate (ATP) and creatine phosphate (CP), calcium homeostasis and the occurrence of oxidative stress, established measuring content and release of reduced and oxidized glutathione. The cytoprotective action of nisoldipine occurs in the absence of negative inotropism and is closely related to the time of administration. Optimal myocardial preservation is achieved when nisoldipine is given before or at the onset of ischaemia. Prophylactic administration of nisoldipine improved the recovery of the developed pressure from 15.9 +/- 1.0 (SE) mmHg to 47.8 +/- 1.9 mmHg, P < 0.01 and reduced the release of CPK from 830 +/- 29 to 229 +/- 27 mU.min-1 g-1 wet wt, P < 0.01. The accumulation of tissue and mitochondrial calcium was reduced from 58 +/- 11 and 49 +/- 9 to 14 +/- 6 and 10 +/- 4 mmol.kg-1 dry wt respectively, P < 0.01. This resulted in a significant (P < 0.01) preservation of all indices of mitochondrial function, allowing a higher recovery of ATP and CP after reperfusion (from 4.1 +/- 0.7 and 10.0 +/- 0.6 to 16.1 +/- 1.0 and 29.9 +/- 0.2 mumol.g-1 dry wt respectively, P < 0.001). Reperfusion-induced myocardial accumulation and release of oxidized glutathione were reduced from 0.493 +/- 0.07 nmol.mg-1 protein and 0.768 +/- 0.063 nmol.min-1 g-1 wet wt to 0.225 +/- 0.07 and 0.157 +/- 0.038 respectively, P < 0.01. Similar data were obtained when nisoldipine was given at the time of ischaemia, while administration 30 min after the onset of ischaemia showed only a trend towards protection. Nisoldipine lost its protective effect when given on reperfusion. A multifactorial analysis of the data suggest that the cardioprotective effect of nisoldipine is related to the maintenance of membrane integrity, possibly since nisoldipine is highly lipophilic.