In order to understand the role of glucose limitations in controlling multicellular tumor spheroid growth, knowledge of the glucose diffusion coefficient is essential. The effective diffusivity of glucose in spheroids of rodent and human tumor cell lines has been determined by measuring the efflux of tritium labeled L-glucose from spheroids with time. When the rapid and irreversible binding of L-glucose in spheroids is properly taken into account, measurements of the efflux of this diffusion tracer from spheroids into label-free medium can be correlated to the diffusion equation in order to obtain the effective glucose diffusivity in spheroids. Such measurements have been made in EMT6/Ro mouse mammary tumor spheroids as well as in spheroids derived from human colon carcinoma cells (HT29, CO112, and WiDr) and from human squamous carcinoma cells (CaSki and A431). EMT6/Ro spheroids have a glucose diffusivity of 1.1 x 10(-6) cm2/s, while glucose diffusion coefficients in the human cell spheroids studied vary from 5.5 x 10(-7) cm2/s to 2.3 x 10(-7) cm2/s. These values are low enough to suggest that significant gradients in glucose concentration may exist in spheroids and tumors. It is thus believed that these glucose diffusivities, as well as their variation with cell line, may have important implications for the role played by glucose in the growth and cellular heterogeneity of spheroids and tumors.