In a previous study of 201 healthy men (1), red cell volume (Vrbc) was measured by a modification of Sterling and Gray's radiochromium method (2), and whole blood and plasma volumes (Vwb and Vpl) were derived from venous hemiatocrits. The influence of factors other than body size on the variance of the data was studied, and standards for predicting normal volumes were derived. The present report describes a similar examination of 101 women. SUBJECTS AND METHODSThe women, all of whom volunteered for study, were actively employed as housewives, laboratory personnel, nurses, or office workers, and most were white (Table I) tResearch Fellow, American Heart Association. prevent venous congestion during sampling. Hematocrit readings were not corrected for plasma trapping, and it was assumed that Vwb = VCr"1 (no allowance was made for a difference between the hematocrit of venous blood and that of the body as a whole) (1). Cell and plasma volumes were calculated by the formulas: Vrhc = VCr`1 X hematocrit; Vpl = VCr5`-Vrbc.The data were treated as outlined in the study of men (1). Regression equations were calculated by the method of least squares to describe the relations between volumes (Vrbc, Vpl, and Vwb) and body measurements (weight, height, weight and height combined, and surface area). The differences between the observed volumes (Vrbc and Vpl) of each woman and the mean volumes found in women of the same height and weight vere calculated by means of Equations 3 and 7, Table II. These differences, or "residuals," wsere used to analyze the influence of factors other than body size On the blood volumes. Table I summarizes the data. In Figure 1, each of the 101 subjects is represented according to her weight and height. In general, the relationship between weight and height was fairly uniform. Notable exceptions were four women weighing over 81 kg, who proved to be 23 to 58 per cent overweight for height and age according to actuarial tables (5). The weight: height ratios of these subjects were more than 2 standard deviations above the mean of the ratios of the other 97 women. For this reason and because the series included so few large women, regression equations describing the relationships of the volumes to body measurements were prepared only for the 97 women weighing less than 74 kg (Table II). RESULTS Volumes in relation to height and weightThe bivariate regression equations described straight lines and the trivariate equations, planes without curvature (see Figures 1 to 4). The location of the regression plane for Vrbc in rela-2182