To simplify the method of plasma volume measurement by Evans blue dye dilution we used, for the first time, the same venous site for injection of dye and collection of samples. In a series of 49 studies the dye decay between 10 and 35 min after injection was highly linear (r = 0.991 +/- 0.01), indicating that contamination of samples is very unlikely. We repeated the measurements after eight weeks in nine patients; the mean difference was 16.4 +/- 19.6 ml, indicating a high degree of reproducibility. We found that extrapolation of the dye decay curve to time zero is required for accurate estimates of plasma volume. There was good agreement between the estimates of plasma volume obtained by extrapolation from only three samples taken at 10, 20 and 30 min after dye injection with the results obtained using all six samples. We also found good agreement between the estimates of plasma volume obtained by using standard curves constructed from four standard dilutions of 1.25, 2.5, 5 and 10 mg/l and those obtained by the use of standard curves constructed from the blank and only one standard dilution of 10 mg/l. We therefore conclude that the Evans blue technique can be simplified with minimal loss of accuracy, by using only one venous site for injection and withdrawal, withdrawing only three samples between 10 and 30 min after injection and using a two point calibration line.