High-resolution 3D seismic P-wave velocity model of Poland (Grad et al., Tectonophysics 666:188–210, 2016) and corrected for paleoclimate heat flow map (Majorowicz and Wybraniec, Int J Earth Sci 100(4):881–887, 2011) gridded to a common mesh are used together with four independent thermal models of the crust and upper mantle to calculate heat flow variation with depth and geotherms. Heat flow at Moho depth are calculated and mapped and both confirm large variability with an elevated mantle heat flow (circa 30–40 mW/m2) in the Paleozoic Platform which is some 10–20 mW/m2 higher than Moho heat flow in the north-eastern and south-eastern Poland which belong to a variety of tectonic terranes (the oldest Precambrian Craton, younger Cadomian, Trans-European Suture Zone, Carpathians). Temperatures calculated for the crust show consistent pattern: higher temperatures beneath the Paleozoic Platform and lower temperatures beneath the Precambrian and Cadomian units. At 10 km depth this difference is about 150 °C, about 300 °C at 20 km depth, and about 400 °C at 50–60 km. Assuming the calculated isotherm 580 °C as Curie temperature the magnetic crust thickness was determined as 5–10 km only beneath the Polish Basin, circa 20 km in Carpathians, circa 30 km in Sudetes, and 35–40 km beneath the Precambrian and Cadomian units. Such a thick magnetic crust results from a great depth of Curie temperature, thick crystalline crust, and thin sediments. Mantle heat flow variability is mainly correlating with measured surface heat flow and influences geotherms. Calculated thermal LAB depth follows patterns of heat flow and Moho heat flow variability through Poland with thinnest lithosphere in the high surface heat flow and high mantle heat flow areas. Comparison of this thermal LAB depth estimates with seismic data based LAB depth shows general coincidences when Precambrian Craton vs Paleozoic Platform are considered along the P4 seismic experiment data model (circa 190 km depth vs some 90 km depth, respectively). However, significant differences exist in many areas and especially for the SE Poland when compared with map for the whole of Poland compiled from other seismic reported data.