A moderately large data set of mostly sedimentary and metamorphic rocks compiled from the literature shows an inverse correlation between thermal conductivity anisotropy and thermal conductivity perpendicular to bedding. The empirical relationship can be represented by a simple two-component end-member mixing model. One end-member is anisotropic (λx/λz~2.5) and has a relatively low thermal conductivity perpendicular to bedding (λz~1 W/m-K); the other end-member is isotropic and has a relatively high thermal conductivity perpendicular to bedding (&ggr;~4 W/m-K). The proposed model may be used to derive a correction for anisotropic effects observed during thermal conductivity measurements on mixtures of randomly oriented aggregates. Although the accuracy and reliability of a universal correction procedure remain to be determined, the proposed correction scheme should enhance our ability to estimate accurately heat flow in sedimentary basins where generally only borehole cuttings are available for thermal conductivity measurements. ¿ American Geophysical Union 1994