Our purpose was 1) to test the hypothesis that in man there is a range of plasma osmolality within which the red cell volume (RCV) and mean corpuscular volume (MCV) remain essentially constant and 2) to determine the upper limit of this range. During a variety of stresses - submaximal and maximal exercise, heat and altitude exposure, +G(z) acceleration, and tilting-changes in plasma osmolality between -1 and +13 mosmol/kg resulted in essentially no change in the regression of percent change in plasma volume (PV) calculated from a change in hematocrit (Hct) on that calculated from a change in Hct + hemoglobin (Hb), i.e., the RCV and MCV were constant. Factors that do not influence RCV are the level of metabolism, heat exposure at rest, and short-term orthostasis (heat-to-foot acceleration). Factors that may influence RCV are exposure to high altitude and long-term orthostatis (head-up tilting). Factors that definitely influence RCV are prior dehydration and extended (>2h) periods of stress. Thus, either Hct or the Hct + Hb equations can be used to calculate percent changes in PV under short-term (<2 h) periods of stress when the change in plasma osmolality is <13 mosmol/kg.