A novel approach based on efficient path-planning algorithms was applied to investigate the influence of substrate access on Burkholderia cepacia lipase enantioselectivity. The system studied was the transesterification of 2-substituted racemic acid derivatives catalysed by B. cepacia lipase. In silico data provided by this approach showed a fair qualitative agreement with experimental results, and hence the potential of this computational method for fast screening of racemates. In addition, a collision detector algorithm used during the pathway searches enabled the rapid identification of amino acid residues hindering the displacement of substrates along the deep, narrow active-site pocket of B. cepacia lipase and thus provided valuable information to guide the molecular engineering of lipase enantioselectivity.