We have carried out two-dimensional (2-D) numerical experiments on the bulk flow of a layer of fluid that is driven in a strike-slip sense by constant velocities applied at its boundaries. The fluid has the (linearized) conventional rheology assumed to apply to lower crust/upper mantle rocks. The temperature dependence of the effective viscosity of the fluid and the shear heating that accompanies deformation have been incorporated into the calculations, as has thermal conduction in an overlying crustal layer. Two end-member boundary conditions have been considered, corresponding to a strong upper crust driving a weaker ductile substrate and a strong ductile layer driving a passive, weak crust. In many cases of practical interest, shear heating is concentrated close to the axial plane of the shear zone for either boundary condition. For these cases, the resulting steady state temperature field is well approximated by a cylindrical heat source embedded in a conductive half-space at a depth corresponding to the top of the fluid layer. This approximation, along with the application of a theoretical result for one-dimensional shear zones, permits us to obtain simple analytical approximations to the thermal effects of 2-D ductile shear zones for a range of assumed rheologies and crustal geotherms, making complex numerical calculations unnecessary. Results are compared with observable effects on heat flux near the San Andreas fault using constraints on the slip distribution across the entire fault system. Ductile shearing in the lower crust or upper mantle can explain the observed increase in surface heat flux southeast of the Mendocino triple junction and match the amplitude of the regional heat flux anomaly in the California Coast Ranges. Because ductile dissipation depends only weakly on slip rate, faults moving only a few millimeters per year can be important heat sources, and the superposition of effects of localized ductile shearing on both currently active and now inactive strands of the San Andreas system can explain the breadth of the heat flux anomaly across central California. ¿ 1998 American Geophysical Union