Transoral laser microsurgery (TLM) and radiotherapy (RT) were stated equal in oncological outcome for early glottic cancer (EGC). But treatment rationale and cordectomy type might vary in TLM. We aimed to provide the clinical experience comparing TLM and RT in EGC. We retrospectively collected data of EGC from a medical center. Primary endpoint was local control (LC), defined as no in situ or invasive carcinoma found in the larynx after definitive treatment. Secondary outcome was total laryngectomy-free rate (TLF). Factors were evaluated with cox regression model. From 2013 to 2020, 58 Tis-T1a and 27 T1b-T2 EGC were included, with 43 underwent TLM and 42 received RT. After medial follow up of 55 months, the 4-year LC was 41.4% vs 78.6% for TLM vs RT, respectively. In multivariate analysis, RT (HR: 0.243, p<0.001) and T1b-T2 (HR: 2.326, p = 0.021) were significantly associated with LC. In subgroup analysis, the difference between treatment modalities was more profound in Tis-T1a subgroup (RT, HR: 0.10, p = 0.002; T1b-T2, RT, HR: 0.488, p = 0.199) and the prognostic value of T classification was less noticed in TLM subgroup (HR: 1.569, p = 0.341; RT, HR: 6.594, p = 0.019), indicating a possible selection bias regarding cordectomy type in Tis-T1a patients receiving TLM. 79.4% of the recurrent events were rEGC. The 4-year TLF rate were 94.9% in TLM and 94.4% in RT. In clinical practice, the LC of TLM might decrease in exchange for a better functional outcome in EGC. Nevertheless, the salvage rate was high that TLM and RT could still yield similar TLF rate.