It is generally assumed that blood is completely mixed while passing through a ventricular chamber. This assumption underlies use of indicator dilution technics for estimation of ventricular volume. In the left ventricle of the exposed canine heart, complete mixing of blood and injected saline during one cycle was rare, especially if the heart rate was in the normal range. Computations of absolute ventricular volume are subject to error if they are based on the assumption that indicator injected into the ventricular cavity becomes uniformly dispersed throughout the ventricular cavity before the end of diastole.