1. This study was designed (i) to assess the effect of S-nitrosoglutathione monoethyl ester (GSNO-MEE), a membrane-permeable analogue of S-nitrosoglutathione (GSNO), on rat isolated heart during cardioplegic ischaemia, and (ii) to monitor the release of nitric oxide (.NO) from GSNO-MEE in intact hearts using endogenous myoglobin as an intracellular .NO trap and the hydrophilic N-methyl glucamine dithiocarbamate-iron (MGD-Fe2+) complex as an extracellular .NO trap. 2. During aerobic perfusion of rat isolated heart with GSNO-MEE (20 mumol 1(-1), there was an increase in cyclic GMP from 105 +/- 11 to 955 +/- 193 pmol g-1 dry wt. (P < 0.05), and a decrease in glycogen content from 119 +/- 3 to 96 +/- 2 mumol g-1 dry wt. (P < 0.05), and glucose-6-phosphate concentration from 258 +/- 22 in control to 185 +/- 17 nmol g-1 dry wt. (P < 0.05). During induction of cardioplegia, GSNO-MEE caused the accumulation of cyclic GMP (100 +/- 6 in control vs. 929 +/- 168 pmol g-1 dry wt. in GSNO-MEE-treated group, P < 0.05), and depletion of glycogen from 117 +/- 3 to 103 +/- 2 mumol g-1 dry wt. (P < 0.05) in myocardial tissue. 3. Inclusion of GSNO-MEE (20 mumol l-1) in the cardioplegic solution improved the recovery of developed pressure (46 +/- 8 vs. 71 +/- 3% of baseline, P < 0.05), and rate-pressure product from 34 +/- 6 to 63 +/- 5% of baseline (P < 0.05), and reduced the diastolic pressure during reperfusion from 61 +/- 7 in control to 35 +/- 5 mmHg (P < 0.05) after 35 min ischaemic arrest. GSH-MEE (20 mumol l-1) in the cardioplegic solution did not elicit the protective effect. 4. During cardioplegic ischaemia, GSNO-MEE (20-200 mumol l-1) induced the formation of nitrosylmyoglobin (MbNO), which was detected by electron spin resonance (ESR) spectroscopy. Inclusion of MGD-Fe2+ (50 mumol l-1 Fe2+ and 500 mumol l-1 MGD) in the cardioplegic solution along with GSNO-MEE yielded an ESR signal characteristic of the MGD-Fe2+ -NO adduct. However, the MGD-Fe2+ trap did not prevent the formation of the intracellular MbNO complex in myocardial tissue. During aerobic reperfusion, denitrosylation of the MbNO complex slowly occurred as shown by the decrease in ESR spectral intensity. GSNO-MEE treatment did not affect ubisemiquinone radical formation during reperfusion. 5. GSNO-MEE (20 microliters l-1) treatment elevated the myocardial cyclic GMP during ischaemia (47 +/- 3 in control vs. 153 +/- 34 pmol g-1 dry wt. after 35 min ischaemia, P < 0.05). The cyclic GMP levels decreased in the control group during ischaemia from 100 +/- 6 after induction of cardioplegia to 47 +/- 3 pmol g-1 dry wt. at the end of ischaemic duration. 6. Glycogen levels were lower in GSNO-MEE (20 mumol l-1)-treated hearts throughout the ischaemic duration (26.7 +/- 3.1 in control vs. 19.7 +/- 2.4 mumol g dry-t wt. in GSNO-MEE-treated group at the end of ischaemic duration), because of rapid depletion of glycogen during induction of cardioplegia. During ischaemia, the amounts of glycogen consumed in both groups were similar. Equivalent amounts of lactate were produced in both groups (148 +/- 4 in control vs. 141 +/- 4 mumol g-1 dry wt. in GSNO-MEE-treated group after 35 min in ischaemia). 7. The mechanism(s) of myocardial protection by GSNO-MEE against ischaemic injury may involve preischaemic glycogen reduction and/or elevated cyclic GMP levels in myocardial tissue during ischaemia.