In this chapter, we describe an in silico approach called CAMP (Co-Culture/Community Analyses for Metabolite Production) to predict two-species microbial communities best suited for producing a desired metabolite. Here, we use genome-scale metabolic models (GEMs) to build microbial communities and constraint-based modeling methods such as flux balance analysis (FBA) to assess and identify suitable communities. Flux variability analysis (FVA) detects the maximum product flux in the communities. The interaction behavior between community members, i.e., mutualism, commensalism, parasitism, and competition, can be deduced based on the variations in the predicted growth rates of the species as monocultures and in co-cultures. In silico community optimization strategies to predict reaction knockouts that improve product flux have also been implemented. CAMP source codes are available from https://github.com/RamanLab/CAMP/tree/master/Protocol .