Spores of Bacillus cereus were heated and recovered in order to investigate the effect of water activity of media on the estimated heat resistance (i.e., the D value) of spores. The water activity (ranging from 0.9 to 1) of the heating medium was first successively controlled with three solutes (glycerol, glucose, and sucrose), while the water activity of the recovery medium was kept near 1. Reciprocally, the water activity of the heating medium was then kept at 1, while the water activity of the recovery medium was controlled from 0.9 to 1 with the same depressors. Lastly, in a third set of experiments, the heating medium and the recovery medium were adjusted to the same activity. As expected, added depressors caused an increase of the heat resistance of spores with a greater efficiency of sucrose with respect to glycerol and glucose. In contrast, when solutes were added to the recovery medium, under an optimal water activity close to 0.98, a decrease of water activity caused a decrease in the estimated D values. This effect was more pronounced when sucrose was used as a depressor instead of glycerol or glucose. When the heating and the recovery media were adjusted to the same water activity, a balancing effect was observed between the protective influence of the solutes during heat treatment and their negative effect during the recovery of injured cells, so that the overall effect of water activity was reduced, with an optimal value near 0.96. The difference between the efficiency of depressors was also less pronounced. It may then be concluded that the overall protective effect of a decrease in water activity is generally overestimated.