The non-uniformity of heating-medium temperature-the most influential process factor-throughout the retort and during the process can reach 10 degrees C. The aim of this work is to facilitate the evaluation of the impact of this variability on safety and quality indexes on the basis of experimental data on temperature histories. The safety of canned foods was estimated by the F-value related to the target micro-organism Clostridium botulinum. The quality was evaluated by the retention of thiamine (vitamin B-1). The employed sensitivity functions show the change in a safety/quality index caused by a unit variation in the heating-medium temperature. The main advantage of this approach is that the parts of the sterilization process which cause considerable changes in safety and quality indexes can be allocated int he space and time domains. The approach is illustrated with a specific case study related to the batch sterilization of in-pact pea puree processed in an over-pressure water-cascading retort.