When sea-level (SL) residents rapidly ascend to high altitude (HA), plasma volume (PV) decreases. A quantitative model for predicting individual % increment PV over the first 7 days at HA has recently been developed from the measurements of % increment PV in 393 HA sojourners. We compared the measured % increment PV with the % increment PV predicted by the model in 17 SL natives living 21 days at HA (4300 m). Fasting hematocrit (Hct), hemoglobin (Hb) and total circulating protein (TCP) concentrations at SL and on days 2, 7, 13, and 19 at HA were used to calculate % increment TCP and % increment PV. Mean [95%CI] measured % increment PV on HA2, 7, 13 and 19 was -2.5 [-8.2, 3.1], -11.0 [-16.6, -5.5], -11.7 [-15.9, -7.4], and -16.8 [-22.2, -11.3], respectively. % increment PV and % increment TCP were positively correlated (P < 0.001) at HA2, 7, 13, and 19 (r(2) = 0.77, 0.88, 0.78, 0.89, respectively). The model overpredicted mean [95% CI] decrease in % increment PV on HA2 (-12.5 [-13.9, -11.1]) and HA7 (-21.5 [-23.9, -19.1]), accurately predicted the mean decrease on HA13 (-14.3, [-20.0, -8.7]), and predicted a mean increase in % increment PV on HA19 (12.4 [-5.0, 29.8]). On HA2, 7, 13, and 19 only 2, 2, 6, and 1, respectively, of 17 individual measures of % increment PV were within 95% CI for predicted % increment PV. These observations indicate that PV responses to HA are largely oncotically mediated, vary considerably among individuals, and available quantitative models require refinement to predict % increment PV exhibited by individual sojourners.