The authors retrospectively analyse 160 patients with primary ocular adnexal lymphoma to determine the accuracy of prognosis based on the Ann Arbor classification and how this compares to the accuracy of prognosis based on retrospectively applying the newer Tumor, Node, Metastasis (TNM) based classification system. The authors conclude that the TNM-based classification is more accurate in predicting prognosis, with bilaterality of disease, positive lymph nodes and / or metastases being important predictors of a poorer prognosis. These factors are recorded in the TNM classification but not the Ann Arbor classification. There is a significant potential confounder, with the treatment for the different stages of ocular adnexal lymphoma, based on the Ann Arbor classification influencing their prognosis, hence, invalidating the retrospective application of the TNM classification. The authors suggest a simplified version of the TNM classification. If patients are split into three groups: 1. unilateral disease, 2. bilateral or positive lymph node or metastasis, 3. bilateral with positive lymph node or metastasis; the following prognosis can be estimated (10 year progression free survival): group 1 – 75%, group 2 – 50%, group 3 – 0%.