A model of perfused cardiac papillary muscle from guinea pig was set up in our lab. The procaine-free St. Thomas' Hospital solution was used as the basic cardioplegic solution. The potassium concentration of the solution was designed by optimization (14.6, 22.8, 28.5, 32.6, 57.8 mmol/L). The papillary muscle was undergone anoxic arrest for 60 min in 32 degrees C. The effect of myocardium preservation was assessed with cardiac action potential, contractility and quantitative analysis of ultrastructure. We concluded that; (1) in this research the proper potassium concentration is at the range of 20.7-26.0 mmol/L and the optimal one is 22.8 mmol/L; (2) despite any deviation from the range, phases II and III of the action potential changes first. [K+] greater than or equal to 32.6 mmol/L slows conductivity, weakens contractility and damages the subcellular structure severely; (3) if [K+] is 57.8 mmol/L, anoxic arrest for 60 minutes in 32 degrees C damages of both structure and function of myocardium is irreversible; (4) following the hyperkalemic cardioplegia, there is a secondary change of cardiac action potential in reperfusion period, which shows longer phase II and shorter phase III, smaller Vmax and APA; (5) reperfusion arrhythmia after anoxic hyperkalemic asystole is likely caused by all kinds of conductive disorders resulting from the secondary changes of action potential; (6) arrest time, AT (Y, s) has the negative relativity to the potassium concentration (X, mmol/L) of cardioplegic solution, which can regress to curve 1/Y = 0.10-1.38/X(14.6 less than or equal to X less than or equal to 57.8). The minimum potassium concentration which can guarantee AT within 30 seconds is 20.7 mmol/L.