The mechanism of twitch prolongation of hypoxic myocardium after reoxygenation was studied before and after interventions that affect cellular cyclic nucleotide levels, subcellular calcium handling, or oxygen-derived free radical production/survival. Right ventricular ferret papillary muscles were subjected to two 20-minute periods of hypoxia, each followed by 1 hour of reoxygenation. The first sequence of hypoxia/reoxygenation was done without intervention. Before the second sequence, the pharmacological agent under study was added to the superfusate or the superfusate calcium concentration was increased from 2.5 to 8 mM. Time from peak to 80% decline in twitch tension was measured in the presence and absence of each intervention immediately before each period of hypoxia and after reoxygenation at maximal twitch prolongation. Interventions that affect Ca2+ flux across the sarcolemma (verapamil and 8 mM [Ca2+]o) or agents that affect oxygen free radical production/survival (dimethyl sulfoxide and allopurinol) did not affect twitch prolongation. Pharmacological agents that increase cyclic AMP levels (forskolin and milrinone) or those that inhibit mitochondrial activity (oligomycin B and ruthenium red) attenuated twitch prolongation. Pharmacological agents that decrease cyclic AMP levels (carbachol) or inhibit function of the sarcoplasmic reticulum (ryanodine) augmented twitch prolongation. The effect of mitochondrial inhibitors on intracellular calcium handling during hypoxia and reoxygenation was examined using muscles loaded with the bioluminescent calcium indicator aequorin. Mitochondrial inhibitors abbreviated the calcium transient and maximal twitch prolongation after hypoxia. We conclude that alterations in sarcoplasmic reticulum and mitochondria calcium handling contribute to the prolonged relaxation seen upon reoxygenation of hypoxic myocardium.