Protein molecules adopt a specific global 3D structure in or- der to carry out their biological function. To achieve this na- tive state a newly formed protein molecule has to fold. The folding process and the final fold are both determined by the sequence of amino acids making up the protein chain. It is not currently possible to predict the conformation of the na- tive state from the amino acid sequence alone and the pro- tein folding process is still not fully understood. We are us- ing L-systems, sets of rewriting rules, to model the folding of protein-like structures. Models of protein folding vary in complexity and the amount of prior knowledge they contain on existing native protein structures. In a previous paper we presented a method of using open L-systems to model the folding of protein-like structures using physics-based rewrit- ing rules. Here we present an L-systems model of pro- tein folding that uses knowledge-based rewriting rules and stochastic L-systems.