We determined the effects of cyclooxygenase-1 (COX-1; SC-560), COX-2 (SC-58125), and inducible nitric oxide synthase (iNOS; 1400W) inhibitors on atorvastatin (ATV)-induced myocardial protection and whether iNOS mediates the ATV-induced increases in COX-2. Sprague-Dawley rats received 10 mg ATV.kg(-1).day(-1) added to drinking water or water alone for 3 days and received intravenous SC-58125, SC-560, 1400W, or vehicle alone. Anesthesia was induced with ketamine and xylazine and maintained with isoflurane. Fifteen minutes after intravenous injection rats underwent 30-min myocardial ischemia followed by 4-h reperfusion [infarct size (IS) protocol], or the hearts were explanted for biochemical analysis and immunoblotting. Left ventricular weight and area at risk (AR) were comparable among groups. ATV reduced IS to 12.7% (SD 3.1) of AR, a reduction of 64% vs. 35.1% (SD 7.6) in the sham-treated group (P < 0.001). SC-58125 and 1400W attenuated the protective effect without affecting IS in the non-ATV-treated rats. ATV increased calcium-independent NOS (iNOS) [11.9 (SD 0.8) vs. 3.9 (SD 0.1) x 1,000 counts/min; P < 0.001] and COX-2 [46.7 (SD 1.1) vs. 6.5 (SD 1.4) pg/ml of 6-keto-PGF(1alpha); P < 0.001] activity. Both SC-58125 and 1400W attenuated this increase. SC-58125 did not affect iNOS activity, whereas 1400W blocked iNOS activity. COX-2 was S-nitrosylated in ATV-treated but not sham-treated rats or rats pretreated with 1400W. COX-2 immunoprecipitated with iNOS but not with endothelial nitric oxide synthase. We conclude that ATV reduced IS by increasing the activity of iNOS and COX-2, iNOS is upstream to COX-2, and iNOS activates COX-2 by S-nitrosylation. These results are consistent with the hypothesis that preconditioning effects are mediated via PG.