A vibrating thermomagnetometer which measures magnetization M continuously at high temperature is used to test a new method of paleointensity determination involving single rather than double heatings. Loss of natural remanent magnetization (NRM) is recorded by the average of 20–25 measurements at the peak temperature T achieved in a zero‐field heating step. Partial thermoremanent magnetization (pTRM) is calculated from the average M after in‐field cooling to room temperature T0. For comparison with data taken at T0, values of M measured at T are multiplied by Ms(T0)/Ms(T), with the aid of the thermomagnetic or Ms(T) curve measured on a chip of the sample. NRM versus pTRM results from 11 heating‐cooling steps on a diabase containing both single‐domain magnetite inclusions in plagioclase and coarser multidomain magnetite grains reproduce the features of Thellier double‐heating paleointensity results for samples from the same site. The NRM – pTRM plot is nonlinear, with convex‐down curvature. For a stringent validation of the single‐heating method, a truly single‐domain sample needs to be tested.