Response surface methodology (Box–Behnken design) was used to evaluate and model effects of three factors (sweetener, low methoxyl (LM) pectin and calcium content) at three levels each, on the overall acceptability of a tropical mixed fruit (pineapple, banana and passion fruit) jelly, determined by 100 consumers. Results showed that the model fit was significant (p = 0.014) and there was satisfactory correlation between actual and fitted values (R2 = 0.940 and adjusted R2 = 0.832). The model presented no significant lack of fit (p = 0.253). Calcium level had a significant effect on overall acceptability, but LM pectin and sweetener levels did not. The statistical model was used to optimise the factors’ levels for highest acceptability, to obtain a jelly that provided less than 12 calories per serving, allowing the product to be labelled as “low calorie”.