Understanding heat flow and its lateral variability is important for petroleum prospecting. We present a methodology and workflow for estimation of radiogenic heat production in the crust and heat flow from geophysical data. The surface heat flow is determined by three main components: heat from Earth’s interior, heat produced in the crust, and heat produced in the sediments. To estimate the crustal heat contribution, we have developed a statistical inversion approach. We first estimate density and susceptibility by inversion of crustal gravity and magnetic anomalies. Then we compute the radiogenic heat production in the crystalline crust by Bayesian rock-physics inversion of density and susceptibility. We demonstrate the proposed methodology with two case examples. In the first example, we compute radiogenic heat production from density and susceptibility measured on rock samples from onshore Norway. The estimated average value of the radiogenic heat production in the rock samples is about 2 μW/m3, with uncertainty of ± 0.5 μW/m3. The second example is from the greater Barents Sea, where we applied a full workflow including a large-scale lithospheric modelling to estimate the mantle heat flow, 3D gravity and magnetic inversion, and rock-physics inversion. The average heat production in the Barents Sea crust is about 1.5 μW/m3. The predicted heat production and heat flow are in good agreement with results presented by other authors.