Eye plaque brachytherapy is a sight-preserving medical procedure in which radioactive sources are reproducibly arranged within a collimating source backing and temporarily sutured to the eye. The procedure was established as an alternative to enucleation or eye removal for the treatment ocular melanoma. The 1987 Collaborative Ocular Melanoma Study (COMS) standardized this treatment technique in a prospective, randomized clinical trial; however, dose calculations were performed using simple assumptions. These assumptions used the point-source dosimetry formalism, omitted dose anisotropy, and ignored the presence of dose perturbing material heterogeneities. Monte Carlo (MC) simulations from the mid-1980s to the present have critically evaluated these assumptions and indicated where improvements in practice could be made. Various investigators have used MC to evaluate plaque design, choice of source radionuclide, X-ray fluorescence from plaque components, plaque material heterogeneity attenuation, and patient tissue heterogeneity attenuation, among others. These publications are reviewed with emphasis on clinically relevant observations. In addition, MC simulations of standard COMS eye plaques using MCNP5 are made and compared to published data using other MC codes. Good agreement is observed between radiation transport codes with differences <2%, for points within the eye globe. These results indicate that historically delivered radiation doses were systematically lower than prescribed doses. Practical considerations for implementing MC calculations in the clinic are introduced, and the need for a heterogeneity-corrected treatment planning system to ensure treatment uniformity across medical centers and for different treatment techniques is discussed.