Crystallographic studies play a major role in current efforts towards protein structure determination. However, despite recent advances in computational tools for molecular modeling and graphics, the task of constructing a protein model from crystallographic data remains complex and time-consuming, requiring extensive expert intervention. This paper describes an approach to automating the process of model construction, where a model is represented as an annotated trace (or partial trace) of the three-dimensional backbone of the structure. Potential models are generated using an evolutionary algorithm, which incorporates multiple fitness functions tailored to different structural levels in the protein. Preliminary experimental results, which demonstrate the viability of the approach, are reported.