The objective of the present work was to develop a formulation for complementary infant and child feeding employing linear programing as a mathematical model for optimization. High lysine/high tryptophan sweet corn pulp in the dehydrated form was used as the main ingredient. The restrictions imposed on the model were nutrient requirements, adequate protein/energy ratio and minimum cost. The formula derived by the computer (FC) matched the amino acid requirements, the protein/energy ratio (NDPCal%) and was rated high in laboratory tests in terms of sensory qualities. The cost determined for this formula was competitive in relation to commercial products used for the same purpose. Formula A, which contained 5% more sweet corn pulp and 10% less whole powdered milk, did not differ in nutritional, sensorial and functional properties from the formula FC and was chosen for the field acceptability trial because of its lower cost. Formula A had protein efficiency ratio and Biological Value similar to casein for the rat but protein digestibility and net protein utilization were statistically lower (p < 0.05) for formula A than for casein. Acceptability tested on children who were 8-18 months of age ranged from 80-90%, average value 87%.