A major issue in the site characterization at Yucca Mountain, Nevada, a potential site for a high-level nuclear waste repository, is the relevance of laboratory-scale measurements on cores to field-scale processes, particularly water flow. Calculation of Philip`s sorptivity parameter using imbibition of water into rock was selected as a simple test to describe hydrologic parameters at both laboratory and field scales and to study effects of sample size and spatial variability. Laboratory-scale imbibition experiments were conducted on two sizes of core from two boreholes drilled in layered nonwelded tuff and fractured welded tuff. Laboratory experiments were compared with field experiments in the boreholes using neutron logs and a field-scale Mariotte system. Measured sorptivity for both sizes of core were virtually identical and both could predict field-scale sorptivity if enough samples were used to account for spatial heterogeneity. Core data was less useful in predicting the neutron log data due to the nature of the neutron probe measurement and difficulties in accounting for effects of the unique system geometry. Mean neutron log values could not predict the field results from the welded borehole due to fractures transmitting but not retaining water, and could not predict laboratory results because of scale and sampling volume differences. The mean neutron log data could, however, approximate the large-scale field results in the nonwelded borehole.