To test one of the assumptions underlying the calculation of stroke volume--namely, that the transthoracic impedance consists of a parallel connection of a tissue impedance and a blood resistance--experiments were carried out on four dogs in which blood was gradually replaced by a stroma free haemoglobin solution, with the purpose of changing the blood resistance while leaving the tissue impedance unchanged. This was accomplished by exchange transfusion in such a way that the volume of the circulating fluid remained constant and the distribution of fluid volume between the fluid compartments was not altered. During the exchange transfusions the mean decrease in resistivity of the circulating fluid was 54%. The packed cell volume and resistivity of every volume of removed circulating fluid were measured. Just before each exchange the real and imaginary parts of the transthoracic impedance were measured. The packed cell volume decreased exponentially with the number of exchanges. This indicates that the circulating fluid volume remained constant during the exchange transfusion. From the packed cell volumes an estimate of the circulating fluid volume was made. Because of the parallel connection all calculations were based on the use of admittance, which is the reciprocal value of impedance. The real and imaginary parts of the transthoracic admittance were calculated from the measured values of the real and imaginary parts of the transthoracic impedance.(ABSTRACT TRUNCATED AT 250 WORDS)