Abstract: Objective To analyze the effects of frequency of knee joint injection therapy, drugs and hospitals on infection rate of knee joint after injection. Methods A total of 121 patients with mean age of (65±11) years (58 males and 63 females) had intra-articular injections within 1 year prior to the hospitalization (from January 2014 to April 2015) in Chinese PLA General hospital. Based on the diagnostic criteria of septic arthritis, 121 patients were divided into infection group (n=25) and control group (n=96). The differences in frequency, drugs (hyaluronic acid or glucocorticoid) and hospitals (grade A tertiary hospital or not) of previous intra-articular injections between two groups were compared. Results Intra-articular injection times of infection group and control group were (4.6±2.2) and (4.1±2.4) with no statistically significant difference (P＞0.05). The proportion of patients undergoing glucocorticoid therapy in infection group and control group were 60% and 34.4% with significant differences in intra-articular injection drugs between two groups (P＜0.05). The proportion of patients undergoing intra-articular injection in non-grade A tertiary hospital in infection group and control group were 92% and 66.7% with significant difference in hospital level (P＜0.05). Conclusion Intraarticular injection with glucocorticoid and injections in non-grade A tertiary hospital are associated with a significant increase in septic arthritis after injections, while the increase of injection times does not lead to an increase in septic arthritis.