Glucose, insulin, potassium (GIK: 300 g glucose + 50 U insulin + 80 mEq KC1/L) was administered to anesthetized dogs as a 30-ml bolus followed by 1.5 ml/kg/h for 2 h. Five populations were studied: control (C, n = 6); 60 min hypothermic arrest both without (I, n = 6) and with pretreatment (I + GIK, n = 6); 60 min hypothermic arrest followed by reperfusion without (R, n = 6) and with pretreatment (R + GIK, n = 6). Glycogen content declined during the ischemic and reperfusion periods whether or not GIK pretreatment was utilized. Glycogen values did not differ significantly among the four groups. GIK pretreatment significantly protected sarcoplasmic reticulum (SR) calcium uptake rates. SR Ca2+ + Mg2+ adenosine triphosphatase (ATPase) activity was unaffected in the I group, depressed in the R group, but protected by GIK pretreatment. Myofibrillar pCa-ATPase activity was significantly depressed in the I group and unaffected by GIK pretreatment. In the R + GIK group, myofibrillar pCa-ATPase activity was identical to controls at all calcium concentrations except for Vmax. In vitro, generation of the superoxide anion by a xanthine-xanthine oxidase system at pH 7.0 significantly depressed both SR calcium uptake and ATPase activity, and this depression was partially reversible by glucose. Generation of the hydroxyl free radical and pH 6.4 significantly depressed calcium uptake but not ATPase activity, and this depression was reversible with glucose + superoxide dismutase. GIK pretreatment exerts a protective effect on the excitation-contraction coupling system during hypothermic global ischemia and reperfusion. Glycogen augmentation after short-term GIK infusion was not significantly different. It is hypothesized that an additional mechanism by which GIK may protect subcellular function is by serving as a scavenger of free radicals generated during the ischemic/reperfusion process.