To evaluate the validity of three equations for estimation of thoracic electrical field size in a new bioimpedance algorithm, stroke volume (SV) as calculated by these equations was compared with that calculated by Doppler echocardiography in 48 healthy volunteers, both lean and obese. When the volume of electrically participating tissue was estimated from body height (modified Sramek) or body height corrected for body habitus (Sramek-Bernstein), there was considerable variation between bioimpedance and Doppler stroke volumes. When the volume of electrically participating tissue was estimated from the actual measurement of the height of the thorax and the circumference at the base of the thorax, the variation in SV differences decreased substantially (Sramek equation), although still considerable for clinical use, and there was no relationship between SV thus obtained and body habitus. Analysis of calculated stroke indices derived by our Doppler echocardiographic standard, as compared with values in the literature, revealed a systematic underestimation. We conclude that the original Sramek equation systematically underestimates SV by 15% to 20%, and the modified Sramek and Sramek-Bernstein equations systematically underestimates SV by 15% to 20%, and the modified Sramek and Sramek-Bernstein equations systematically overestimate SV in females by about 15%, but provide SV values in males in the predicted range. Further studies on the current assumption that the electrical field size is a truncated cone may improve precision of the bioimpedance method.