Determinations of absolute paleointensity ar often hampered by magnetomineralogical changes produced within samples of lava flows during successive heatings. Such change which directly affect the capacity for a sample to acquire a partial thermoremanent magnetization (PTRM) are reflected by negative pTRM checks, a deviation between the original pTRM and a repeated measurement after heating at a higher temperature. Because the deviations indicated by the PTRM checks can provide a direct estimate of the effects to the pTRMs, we suggest that they can be used to correct pTRMs. The correction method we present requires additional measurements, in particular pTRM checks performed after each pTRM and specific diagrams to detect acquisition of chemical remanence. Performing multiple successive pTRM checks can also provide information on the blocking temperatures of the alteration product. These experiments conducted on lava flows from different localities show that the magnetomineralogical changes mainly involved grains with blocking temperatures lower than the last heating step. The corrections have been tested on recent and historic lava flows from several localities. We have also compared results on samples from the same lava flows that were heated in air and in vacuum. The mean paleointensity values were found to lie within less than 10% from the expected field intensity at each site. The corrections resulted in the recovery of reliable paleointensity determinations for about twice as many specimens heated in vacuum and 65% additional data for specimens heated in air. Successive heatings at the same temperature in air are often accompanied by a time-dependent oxidation of the natural remanent magnetization which is responsible for paleointensity values lower than the expected field, a problem that is not related to the corrections. We conclude that corrections can be used with fairly high confidence to the pTRMs performed in vacuum and in air. ¿ American Geophysical Union 1996