We compared responses of turtle heart at 20 degrees C to an anoxic lactic acidosis solution (LA) containing 35 mM lactic acid in an otherwise normal turtle Ringers equilibrated with 3% CO2/97% N2 at pH 7.0) to a solution simulating in vivo anoxic acidosis (VA), with elevated concentrations of lactate, Ca2+, Mg2+, and K+, and decreased Cl-, equilibrated with 10.8% CO2/89.2% N2 at pH 7.0. We examined mechanical properties on cardiac muscle strips and determined intracellular pH (pHi) and high energy phosphates on perfused hearts using 31P-NMR. Maximum active force (Fmax) and the maximum rate of force development (dF/dtmax) of muscle strips were significantly higher during VA than during LA superfusion. An elevation of Ca2+ alone (to 6 mM) in LA significantly increased both Fmax and dF/dtmax but the effects diminished toward the end of the exposure; however, hypercapnic anoxic lactic acidosis (addition of 20 mM HCO3- to LA, equilibrated with 10.8% CO2/89.2% N2, pH 7.0) did not significantly affect Fmax or dF/dtmax. During VA perfusion, pHi (6.73 +/- 0.01) was significantly higher than that during LA perfusion (pHi 6.69 +/- 0.013), but the difference is probably too small to have physiological significance. ATP, creatine phosphate, and inorganic phosphate were not significantly different in the two anoxic solutions. We conclude that the reduction of cardiac mechanical function in vivo is minimized by the integrated effects of changes of ionic concentrations, but the observed changes in Ca2+ and pHi cannot fully explain the effect.