Previous experiments carried out on several carbonization phases at laboratory and pilot scales have shown a relationship between the migration of tar, evidenced by extraction with N-methyl-2-pyrrolidinone (NMP), and the coking pressure. In the present study, complementary experiments using similar analytical techniques confirm that the impregnation of non-transformed coal by recondensed tars is related to the wall pressure that develops during pyrolysis. The results obtained on partially coked coal charges in a laboratory oven permit the simulation of the evolution of impregnation versus the blending ratio of binary blends. Compared with pilot-scale measurements of the wall pressure on various blends, this evolution of impregnation appears to be directly related to the wall pressure. On this basis, a predictive laboratory test of wall pressure evaluation through solvent extraction is proposed.