Background: Currently, endoscopic resection of early colorectal cancer (ECC) defined as carcinoma with invasion limited to the mucosa (Tis), and submucosa (T1) is possible due to advances in instrumentation. However, when tumor invades submucosal layer, lymph node dissemination is present in 16.2% of cases, requiring additional surgery and limfadenectomy. Risk factors for lymph node dissemination and independent for relapse have been previously described in literature. Methods: We performed a retrospective analysis of all patients with colorectal T1 tumors, treated at our center with endoscopic resection and some with additional surgery between 2006 and 2017. Stadistical analysis was perfomed with IBM SPSS Statistics 24.0. Results: 159 patients (p) were treated with endoscopic resection, 56.6% (90p) underwent additional surgery. Mean age was 68.74 years and 69.9% (111p) were male. Endoscopic resection: negative margins 87.6%, vascular 3.1%, lymphatic 2.5% and perineural invasion 3.8%, high degree of histological differentiation 1.3%. Surgical resection: negative margins 100%, lymph node spread 8.8%. In a mean follow-up of 23.36 months since endoscopic treatment, 13 patients had relapsed. Risk of relapse did not differ between patients who received additional surgery and those who only underwent endoscopic resection (p = 0.791). On the other hand, lymph node dissemination (p = 0.007) and a positive margin (p = 0.01) were independent risk factors for risk of relapse. Vascular, lymphatic and perineural invasion, nor degree of histological differentation were stadistically significant. However, there was a positive association between lymph node dissemination and lymphatic (p = 0.07), vascular (p = 0.007) and perineural (p = 0.001) invasion and also degree of histological differentiation (p = 0.000). Conclusions: In our study, relapse rate was under 10% in eleven years. The only independent risk factors for relapse were a positive margin and lymph node dissemination. Perineural, vascular and lymphatic invasion obtained from polypectomy sample, could infer the probability of positive lymph nodes and indirectly be an indicator for risk of relapse.