Six most commonly known models for predicting thawing time of food: Nagaoka's et al., Cleland's et al., Calvelo's, Pham's and Piotrovich were compared when testing Tylose MH 1000 test substance, ground beef, and potato, and relative errors, regression and variance. The Cleland's et al. method, disregarding equivalent heat transfer dimensionality (EHTD) and mean conducting path (MCP) coefficient, was proved to be the best for predicting thawing time. The inclusion of EHTD and MCP to the computations by the Clelaçd's et al. method did not affect the results statistically significantly. The models of Piotrovich, Calvelo and Nagaoka et al. produced the results statistically different from real thawing times.