Background: MRI predicted distance of mesorectal extension (mrDME) in rectal cancer is one of the independent risk factors for recurrence and poor overall survival. In T3 rectal cancer, if no lymph node or distant metastasis is seen, the selection of optimum treatment is based on the distance of mesorectal extension. Therefore, it is very crucial to investigate the reproducibility of DME in T3 rectal cancer. Objectives: To investigate the reproducibility of the distance of mesorectal extension by tumor invasion in T3 stage rectal cancer by evaluating sub-stages T3a, T3b T3c and T3d individually versus T3a, T3b (T3ab) and T3c, T3d (T3cd) combined together using MRI. Methods: From July 2014 to December 2015, 188 patients with surgically and histologically confirmed T3 rectal cancer who underwent preoperative MRI were enrolled into this study. Two blinded radiologists evaluated the maximum distance of mesorectal extension (mrDME) in T2 weighted image in MRI. The study population was sub classified into T3a (< 1 mm), T3b (1 -5 mm), T3c (5 -15 mm) and T3d (> 15 mm) according to the distance of mesorectal extension by tumor invasion. The inter-observer and intra-observer data were then assessed using Kappa analysis (k); intraclass correlation coefficients (ICC) were also analyzed subsequently. Results: Difference in the value of inter/intra observer kappa (k), and inter/intra-observer intraclass coefficients (ICC) between the two groups was very distinct. In the individual group (T3a, T3b, T3c and T3d), the inter-observer and intra-observer (k) for the mrDME was 0.700 and 0.718 respectively; the inter-observer and intra-observer ICC was 0.772 and 0.786 respectively. In the combined group (T3ab and T3cd), the inter-observer and intra-observer kappa (k) for the mrDME was 0.819 and 0.883 respectively; the inter-observer and intra-observer ICC was 0.829 and 0.796 respectively. Conclusions: There was a distinct increase in the kappa (K) and intraclass coefficient (ICC) value in the combined group compared with the individual group. This high reproducibility result suggested that it is more reliable to measure T3ab and T3cd combined together than individually. This finding can play a crucial role in the management of rectal cancer and clinical decision making for non-expert radiologists in non-academic setting.