The effects of cultivation medium compositions including tapioca, ®shmeal, CaCO 3 and (NH 4 ) 2 SO 4 for the growth of Bacillus thuringiensis YMB 96-1988 were accessed by using response surface methodology (RSM). The two-level (2 4±1 ) fractional factorial designs (FFD) which involve two concentrations of each nutrient, and the paths of steepest ascent were effective in searching for the major factors of the bacteria growth. This allows the ®tting of a ®rst order linear model to the data. In this study, supplementary CaCO 3 showed a negative effect on the spore production based on the ®rst order regression coef®cients derived from SAS programme. Subsequently, a 2 3 central composite design (CCD) was used for allocation of treatment combinations. Preliminary studies showed that tapioca and ®shmeal is believed to be the major factors for the growth of B. thuringiensis. Estimated optimum compositions for the production of spores by B. thuringiensis are as follows: tapioca, 5.01%; ®shmeal 5.86%; (NH 4 ) 2 SO 4 0.06% and resulted in a maximum spore count of 8.56´10 8 /ml was obtained. This value is close to the 8.35´10 8 /ml spore density as counted from actual experimental observations.