The effect of cardiopulmonary bypass (CPB) on myocardial extravascular water (MEW) was evaluated with crystalloid and colloid hemodilution. Heart water was measured gravimetrically and by the double-indicator and thermal methods. CPB without hemodilution resulted in a 5.7 per cent increase in the wet : dry weight ratio of the left ventricle obtained by desiccation to stable weight. CPB with colloid hemodilution to a hematocrit of 10.7 +/- 0.4 per cent resulted in a 5.4 per cent increase in the wet:dry weight ratio. Crystalloid hemodilution to a hematocrit of 9.5 +/- 0.8 per cent resulted in a marked increase in myocardial water with a wet:dry weight ratio 30.3 per cent greater than the controls. Hypothermic (22 degrees C.) crystalloid hemodilution resulted in a 37.4 per cent increase in the wet:dry weight ratio. MEW was also measured by the double-indicator method with Evans blue dye and tritiated water. This method measured 85 per cent of the gravimetrically measured water. Although it indicated the increase in heart water in the crystalloid group, it proved less reliable in the measurement of MEW in this dynamic situation. The thermal heart water was also measured with an impedance and thermistor-bearing catheter similar to that used to measure thermal lung water. This proved ineffective in measuring heart water. Colloid hemodilution was thus found to prevent the development of myocardial edema which occurred with crystalloid hemodilution (p less than 0.01) with and without hypothermia. These findings support the addition of colloid to the hemodilution prime used for cardiopulmonary bypass.