Pasteurized process cheese spreads with various levels of sodium chloride, disodium phosphate, moisture and pH were challenged with spores of Clostridium botulinum types A and B. Response surface methodology was used to design experiments that would yield maximum results with the minimum number of trials. Supplemental experiments were added to further clarify the response and to examine combinations of special interest. A total of 304 treatment combinations (batches) was incubated at 30°C, and five samples from each batch were taken at predetermined intervals up to 42 wk of incubation and tested for botulinal toxin. Sodium chloride and disodium phosphate inhibited botulinal toxin production with similar effectiveness. The inhibitory effect of low pH (<5.7) and low moisture (<54%) levels on botulinal toxin production was as expected, i.e., as either pH or moisture went up, it was necessary to increase sodium chloride and/or phosphate concentrations to compensate. Differences in water activity between cheese spreads with different compositions were observed but they were too small to use for controlling the properties of the products, e.g., a range of 9% in moisture level (51 to 60%) produced only 0.022 variation in water activity. Combinations of the above factors were developed for safe pasteurized process cheese spreads containing up to 60% moisture.