OBJECTIVE: This network meta-analysis (NMA) aimed to compare the efficacy of various physical modalities in alleviating pain, stiffness, and functional impairment in patients with knee osteoarthritis (KOA). METHODS: In accordance with PRISMA-P guidelines, we systematically searched nine databases(CNKI, VIP Database, Wanfang Database, SinoMed, PubMed, Embase, CINAHL, Web of Science, and the Cochrane Library) from inception to October 2024 to identify randomized controlled trials (RCTs) evaluating physical therapies for KOA. The interventions assessed included electrical stimulation therapy (EST), low-level light therapy (LLLT), thermotherapy (TT), cryotherapy (CT), and extracorporeal shock wave therapy (ESWT), with resistance and range of motion exercises (RRE) serving as comparators. Outcome measures comprised the Visual Analog Scale (VAS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and 6-minute walk test (6 MWT). Bayesian network meta-analyses and pairwise meta-analyses were performed using Stata 17.0 and R 4.4.1 software. RESULTS: 32 RCTs involving 2,078 participants were included. LLLT demonstrated the highest efficacy for pain reduction (VAS: MD=-3.32, 95% CI:-3.82 to-0.75; WOMAC pain: MD=-3.74, 95% CI:-6.68 to-0.72) and joint function improvement (SUCRA = 79.8). ESWT ranked second for pain relief (VAS: MD=-1.31, 95% CI:-2.42 to-0.16) and mobility enhancement (6 MWT: SUCRA = 71.5), while TT showed superior efficacy in reducing stiffness (WOMAC stiffness: MD=-2.09, 95%CI:-3.06 to-0.94; SUCRA = 98.1). In contrast, ultrasonic therapy (UT) did not provide significant benefits. CONCLUSIONS: The findings suggest that LLLT and ESWT may be optimal for pain relief and functional improvement in patients with KOA, whereas TT appears to be the most effective in reducing stiffness. Optimal dosing parameters of these physical modalities are crucial for maximizing clinical benefits. Clinicians should individualize treatment strategies based on patient-specific factors. Future large-scale RCTs are warranted to validate these protocols and address the heterogeneity of existing evidence. CLINICAL TRIAL NUMBER: Not applicable.