The purpose of this study was to examine the discriminant and predictive validity of grip strength measurements obtained dynamometrically from patients in acute rehabilitation. Retrospective chart reviews were performed to retrieve information on admission grip strength. That strength was compared to published norms to establish discriminant construct validity. Predictive validity was investigated by analyzing the relationship of grip strength with rehabilitation length of stay (LOS) and discharge self-care function (Functional Independence Measure scores). Wilcoxon signed ranks tests showed that the patients' grip strength was significantly less than normal on both their left (Z=-9.205, p<.001) and right (Z=-10.729, p<.001) sides. Grip strength was a significant but weak predictor of both LOS (Spearman rho=-0.277 to -0.352) and discharge function (Spearman rho=0.360 to 0.382) Multiple regression revealed that grip strength added nothing further to the prediction of LOS or discharge self-care function once admission self-care function was taken into account. This study provides strong support for the discriminant validity of grip strength measurements but only weak support for their predictive validity in acute rehabilitation.