The potential risk of C. botulinum growth in fresh fish stored under modified atmospheres remains unclear. Few studies have identified qualitatively certain conditions leading to toxigenesis. This paper is the First of a series attempting to quantify the effect of a variety of parameters on the probability (P) of toxigenesis by one spore in fish. The factorial design experiments included red snapper tissue homogenate inoculated with a pool of nonproteolytic spores (5 type E, 4 type B and 4 type F strains) at 7 levels (104−10−2 per 3 g sample) and incubated at 4, 8, 12, 17 and 30°C under 3 modified atmospheres (vacuum, 100% CO2 and 70% CO2+30% air) for up to 21 days. At the 100 spore/sample level the earliest time to detect toxin production at 4, 8, 12, 17 and 30°C under all modified atmospheres was > 21, 12, 9, 6 and 2 days, respectively. At the 101 spores/sample level the earliest times for the same temperatures were > 21, 9, 6, 3–6 and 1–2 days, respectively. The probability of toxigenesis was affected significantly (P < 0.005) by temperature, storage time, atmosphere×temperature, and temperature×time but not by atmosphere (P > 0.1). Using linear and logistic regression models, equations were derived which can predict the P of 1 spore initiating growth and toxigenesis by a particular day and at a particular temperature of storage. Studies involving other fish substrates are in progress.