AIM This study was aimed to compare perioperative and oncological outcomes of mid/high locally advanced midrectal cancer (LARC) treated with neoadjuvant chemoradiotherapy (NCRT) vs. surgery alone, and to identify risk factors for local recurrence in mid/high LARC. METHOD A total of 471 mid/high LARC patients treated with surgery alone or NCRT (50.4 Gy in 28 fractions) with concurrent FOLFOX/XELOX followed by TME in 6-8 weeks from 2008 to 2014 were matched 1:1 by using propensity score analysis. Perioperative and survival outcome was compared between groups. Multivariate analyze was performed to identify risk factors for local recurrence. RESULTS Two hundred and two patients were matched for the analysis. Postoperative morbidity was similar between groups. With a mean follow-up of 57 months, the 5-year overall survival (NCRT vs. surgery alone: 80.4% vs. 81.4%; p = 0.978), 5-year local recurrence rates (3.1% vs. 5%; p = 0.467), and 5-year distant metastasis rates (29.5% vs. 23.7%; p = 0.140) were similar between two groups. Cox regression analysis showed that the circumferential resection margin (CRM) involvement (OR = 5.205, p = 0.005) was the only risk factor for local recurrence in mid/high LARC patients. CONCLUSION In matched cohorts of mid/high LARC patients, surgery alone provided comparable oncological outcome, when compared with NCRT. CRM involvement was the only risk factor for local recurrence in mid/high rectal cancer. NCRT may not be always needed in mid/high LARC. A threatened CRM could be diagnosed up front and prevented by selecting CRT for these patients.