Background: Pre-operative CRT followed by total mesorectal excision (TME) is nowadays the standard of care for patient with LARC (cT3-T4N0 or cTxN+). Currently, pathologic complete response occurs in +/- 15% after CRT. Colorectal cancer T-cell infiltration is a strong prognostic factor for survival after primary tumor resection. Our aim was to determine whether T-cell infiltration in PTB could be predictive of tumor response and relapse after CRT + TME. Methods: Between 1999 and 2012, patients with LARC who underwent CRT + TME and with available clinical follow-up and PTB (with sufficient tumor cells density) were identified at the Cliniques universitaires St-Luc. The density of CD3 (T cells) and CD8 (cytotoxic) was quantified on immunostained PTB slides and analyzed with a dedicated image analysis software on whole-slide imaging. Comparisons were made using the Wilcoxon-Mann-Whitney test. Cumulative disease-free survival (DFS) was performed using the Kaplan-Meier estimator and compared by log-rank tests. Cox regression we used for uni- and multi-variate analysis. P value of less than 0.05 was considered statistically significant. Results: 154 patients (sex ratio M/F 1.8; mean age 65 years-old; upper (20%), mid (29%) and low rectum (51%), synchronous metastases (11%)) were analyzed. High CD3 and CD8 PTB densities were significantly associated with a higher pathological response (Dworak 3-4) and lower ypTNM stage after CRT +TME (p < 0,05). Higher CD3 and CD8 PTB densities were associated with higher patient DFS (CD3: HR = 2,30 (CI95%:1,15-4,59) p = 0,02; CD8: HR = 1,95 (CI95%: 1,01-3,75) p = 0,04). These results were confirmed in uni and multivariate analysis. CD3 and CD8 PTB densities added to pathological response (ypTNM/Dworak) but also clinical response (ycTNM) after CRT + TME increases significantly the accuracy prediction of tumor relapse. Conclusions: Pretherapeutic T-cell infiltration of LARC is predictive of tumor response and relapse after CRT +TME. This biomarker could be helpful for patient treatment decision. It must be validated in larger patient cohorts.