Background and aims Endoscopic submucosal dissection (ESD) as an advanced endoscopic procedure can be considered for the removal of colorectal lesions with high suspicion of limited submucosal invasion or cannot be optimally removed by snare-based techniques. We aimed to analyze the clinical outcomes of ESD for colorectal neoplasms in our hospital. Methods We retrospectively enrolled 230 patients with 244 colorectal neoplasms who received ESD procedures from April 2012 to October 2020 at Kaohsiung Chang Gung Memorial Hospital. Clinicopathological data were collected by chart review. We also recorded ESD-related complications and clinical outcomes. Results The average age was 64 years old, with a mean follow-up time of 22.59 months. There was a loss of follow-up in 34 lesions. Most lesions were lateral spreading tumors of the non-granular type. The average ESD time was 51.9 minutes. Nine cases (3.7%) had procedure-related complications, including two intra-procedure perforations (0.8%) and seven delayed bleeding (2.9%) without procedure-related mortality. 241 lesions (98.8%) achieved en-bloc resection, while 207 lesions (84.8%) achieved R0 resection. Most lesions were tubulo-(villous) adenoma. Malignancy included 35 adenocarcinomas and 5 neuroendocrine tumors. No local recurrence was developed during follow-up. Multivariate analysis for long ESD time revealed significance in size ≥ 10 cm2 and endoscopist’s experience < 3 years. Pre-ESD endoscopic ultrasound revealed good prediction in discrimination of mucosal (sensitivity: 0.90) and submucosal lesion (specificity: 0.67). Conclusions ESD for colorectal neoplasms is an effective and safe technique. Size ≥ 10 cm2 and endoscopist’s experience < 3 years were significantly associated with long procedure time. Pre-ESD EUS provided a good prediction for colorectal neoplasms in invasion depth.