BACKGROUND: The aim of this study was to identify the indications and oncological outcomes of selective lateral lymph node dissection (sLLND) in rectal cancer patients. METHODS: A retrospective study was conducted on consecutive patients with rectal cancer who had standard total mesorectal excision and sLLND at our institution. Clinicopathological characteristics and oncological outcomes were analyzed. We performed subgroup analysis and multivariate analysis based on patients with or without preoperative chemoradiotherapy to identify the related risk factors. RESULTS: A total of 77 consecutive patients with TME and sLLND were included. Twenty-two (28.6%) patients with pathological positive lateral lymph nodes metastasis (LLNM) were identified. Forty-seven (61%) patients accepted neoadjuvant chemoradiotherapy (nCRT). The pretreatment maximum short-axis diameters of LLN (≥ 8 mm) were the independent risk factors for LLNM among patients with LLN ≥ 5 mm. Lymph node metastasis were significantly higher in patients with pretreatment LLN ≥ 8 mm than in patients with LLN 5-8 mm (63% vs. 10%, p < 0.001). The receiver operating curve analysis suggested that the optimal cutoff value of LLN short-axis diameter for predicting LLNM was 8 mm. At a median follow-up of 42 months (range 6-140 months) 3 (3.9%) patients with lateral pelvic recurrence were observed. The 3-year cumulative overall survival in patients with LLNM and patients without LLNM was 76.7% and 89.8%, respectively (p = 0.01). The 3-year cumulative disease-free survival was 53.6% in patients with LLNM and 88.3% in patients without LLNM (p = 0.008). CONCLUSION: Patients with LLNM had a worse prognosis. The pretreatment maximum short-axis diameter of LLN (≥ 8 mm) should be considered as an indication for sLLND.