Soluble epoxide hydrolases (sEHs) catalyze the hydrolysis of epoxides to their corresponding vicinal diols. One property of a number of these enzymes is that they can catalyze the hydrolysis of some racemic substrates in an enantioconvergent one-enzyme fashion. Here, we have used the dispersion-corrected B3LYP-D3 density functional theory method to investigate the enantioconvergent conversion of styrene oxide (SO) by sEH from Solanum tuberosum (StEH1). A large cluster model of the active site, consisting of 279 atoms, is designed on the basis of the X-ray crystal structure of StEH1 in complex with the competitive inhibitor valpromide. Different substrate orientations of the two enantiomers of SO are examined, and the full reaction mechanisms for epoxide opening at the two carbons are calculated, including both the alkylation and hydrolysis half-reactions. The calculated overall reaction energy profiles show that the rate-determining step is associated with the dissociation of the covalent intermediate, which is the second step of the hydrolysis half-reaction. The calculations reproduce the experimentally observed regioselectivities for the two enantiomers of the substrate, in that both (S)-SO and (R)-SO are calculated to yield the same (R)-diol product. The obtained energy profiles indicate that the transition states for both the alkylation and hydrolysis half-reactions have to be taken into account in order to understand the stereochemical outcome of the reaction. The transition state structures are analyzed in detail, and several factors that contribute to the selectivity control are identified. In addition, the mechanistic scenario in which the active site His300 residue is in the protonated form is also considered and the implications on the energies and enantioselection are discussed. The current calculations demonstrate the applicability of the quantum chemical cluster methodology in reproducing and rationalizing experimental enantioselectivities, lending further support to its usefulness as a tool in asymmetric biocatalysis. The results presented here can be helpful in the rational engineering of sEHs to obtain variants with refined biocatalytic properties.