Insect diets are complex mixtures of vitamins, salts, preservatives, and nutrients. To determine the effect of varying doses of multiple components, the traditional approach requires large factorial experiments resulting in very large numbers of treatment combinations and multiple interaction terms that are difficult or impossible to interpret simultaneously. An insect diet can be conceptualized as an n-dimensional space requiring multi-variate analysis. Software packages that take advantage of the increased computing capability of modern microcomputers now make it possible to design and execute highly efficient experimental designs that systematicaly sample through multidimensional space, identify key drivers for specific response variables, and generate mathematical equations that describe multiple response variables. The result is a set of equations that describe reponses and allow researchers to choose the most desirable outcome. Our immediate interest was to optimize a commercial diet for the tropical weevil Diaprepes abbreviatus, a pest of citrus, ornamentals and other crops in the Caribbean, Florida and California. We also present the application of response surface methodology to the problem of diet optimization and illustrate how this approach can produce insect diets optimized for any measurable criterion desired by the researcher with a minimum of experimentation.