Abstract
Plasma volume has been estimated in 10 human subjects using Coomassie-blue and I131 radioiodinated human serum albumin dilution methods simultaneously. Three different methods of correction used by previous workers to overcome the error due to early dye loss were applied. Satisfactory agreement with the established radioiodinated albumin method was only obtained by extrapolation of the semilogarithmic plot of Coomassie-blue plasma dye concentration between five and 10 minutes to the time of injection. The significance of the controversial Evans blue "mixing curve" is discussed. An analogous phase in the Coomassie-blue disappearance slope is considered to be due to initial rapid loss of dye from the circulation rather than to the process of mixing. It is shown that Coomassie-blue fulfils the criteria listed in the discussion for plasma volume estimation.