Objective: The aim of the present systematic review and meta-analysis is to know, based on the available randomized controlled trials, if the non-surgical and non-pharmacological interventions commonly used for knee osteoarthritis (OA) patients are effective and which are the most effective ones. Material and Methods: RCTs were identified through electronic databases respecting the following terms to guide the search strategy: PICO (Patients – Humans with knee OA; Intervention – Non-surgical and non-pharmacological interventions; Comparison – Pharmacological, surgical, placebo, no intervention, or other non-pharmacological/non-surgical interventions; Outcomes – Pain, physical function and patient global assessment). The methodological quality of the selected publications was evaluated using the PEDro and GRADE scales. Additionally, a meta-analysis was performed using the RevMan. Only studies with similar control group, population characteristics, outcomes, instruments and follow-up, were compared in each analysis. Results: Initially, 52 RCTs emerge however, after methodological analysis, only 39 had sufficient quality to be included. From those, only 5 studies meet the meta-analysis criteria. Exercise (especially resistance training) had the best positive effects on knee OA patients. Pulsed Electromagnetic Fields and Moxibustion showed to be the most promising interventions from the others. Balance Training, Diet, Diathermy, Hydrotherapy, High Level Laser Therapy, Interferential Current, Mudpack, Neuromuscular Electrical Stimulation, Musculoskeletal Manipulations, Shock Wave Therapy, Focal Muscle Vibration, stood out, however more studies are needed to fully recommend their use. Other interventions did not show to be effective or the results obtained were heterogeneous. Conclusions: Exercise is the best intervention for knee OA patients. Pulsed Electromagnetic Fields and Moxibustion showed to be the most promising interventions from the others options available.