Background::There is limited evidence of the effects of local anesthesia (LA) on outcomes of non-surgical periodontal treatment (NSPT), in particular among the Chinese. This retrospective cohort study aimed to evaluate the effects of LA on short-term treatment outcomes of NSPT and to determine under what circumstances LA should be prescribed to improve these outcomes.Methods::Data from periodontal examinations of 3980 patients were used. The data were from 3-month re-evaluation records of an electronic periodontal charting record system in the Department of Periodontology of Peking University School and Hospital of Stomatology from June 2008 to January 2015. Descriptive analyses included changes in probing depth (PD) and the Mazza bleeding index (BI). Two-level (patient and tooth) logistic regression models and three-level (patient, tooth, and site) linear regression models were constructed to analyze the influence of LA on PD for all teeth/sites and teeth/sites with an initial PD ≥ 5 mm. Decreases in PD and BI at sites under LA using the initial PD were also compared.Results::A significantly higher mean decrease in PD after NSPT was found in the LA group than in the no local anesthesia (NLA) group (0.98 n vs. 0.54 mm, n t = 24.12, n P 1 and BI > 2) for all teeth (16.7% n vs. 13.8%, n t=3.75, n P < 0.001; 34.7% n vs. 28.1%, n t = 6.73, n P < 0.001) and PD for teeth with PD ≥ 5 mm (32.3% n vs. 17.3%, n t = 28.48, n P < 0.001). The difference in PD between the LA and NLA groups increased as the initial PD increased. The difference between the two groups was 0.12 to 0.22 mm for sites with a baseline PD < 7 mm; however, it increased to 0.41 to 1.37 mm for sites with a baseline PD ≥ 7 mm.n Conclusions::LA improved the decrease in PD after NSPT. Root debridement at sites with initial PD ≥ 7 mm should be performed under routine LA.