Purpose: Conventional treatment planning for episcleral plaque brachytherapy entails creating a two‐dimensional (2D) isodose plan based on the tumor dimensions and distances to critical structures as defined by ultrasound and fundus photography. The purpose of this study was to compare 2D‐based metrics of the tumor and normal eye anatomy with magnetic resonance (MR)‐based metrics. Methods: Thirty‐five patients with ocular melanoma were prescribed to receive 85 Gy over 96 hours using episcleral plaque brachytherapy with iodine‐125. The plaque size and seed source strength were determined from the tumor characteristics as defined by ultrasound and fundus photography. The patients were imaged on a 1.5‐Tesla MR unit using T1‐and T2‐weighted techniques. Images were exported to a brachytherapy planning system for contouring of tumor and normal eye anatomy, and identifying points of interest for critical structures, e.g., fovea and optic disk. The tumor apex height, basal dimensions, distances between tumor edge to fovea and to optic disk were averaged over all patients, and compared between 2D‐and MR‐based methods using statistical Signed test analysis. Results: There were no statistically significant differences between 2D‐and MR‐based methods for mean tumor apex height, basal dimensions, and distance to optic disk and to fovea. Four patients presented with tumors located anteriorly in the ciliary body. This tumor location proved to be challenging in estimating distances to optic disk and fovea based on 2D methods. These distances were measurable on MR regardless of tumor location. Conclusion: We have shown that tumor dimensions and distances measured using 2D‐based methods can be reproducibly defined on MR. Future work will evaluate isodose plans on the MR images and compare them to 2D‐based isodose plans. Additional work will include investigating different MR imaging techniques to enhance soft‐tissue visualization.