Drying of microbial cells for storage purposes is a widely practised technology. A new statistical model is proposed for survivability of microbial cells during convective air drying, where mortality is predicted by hazard functions and acceleration factors applied to temperature, moisture content and drying rate variables. The model is based on experimental survivability data generated from drying Penicillium bilaiae conidia at temperatures from 20 to 60 degrees C and air relative humidities from 3 to 75%. Experimental data using "one at a time" variable manipulation was used to obtain five model parameters. Despite several simplifications to reduce the number of variable cross-dependencies, the model was successfully validated using combined stress trials with maximum deviations of +/- 15%.