Therapeutic exercises in osteoarthritis (OA) are therapeutic approach with proven efficacy. However, the different disease forms, the lack of established consensus and system for patient education limit the widespread use of therapeutic exercises in clinical practice. The mechanisms of action of therapeutic exercise in OA include the following components e.g. neuromuscular (improvement of proprioception, muscle strength and joint stability), intra-articular (prevention of cartilage degeneration, antiinflammatory effect, improvement of the quality of the joint fluid), periarticular, general health with reduction of cardiovascular risk and psychological components. The main exercises, which are used in OA patients, are stretching exercises, range of motion exercises, analytic exercises for muscle strengthening (isometric and isotonic) and aerobic exercises (walking, cycling, swimming, etc.). The recommended exercise programmes have to be individualized according to patient age, severity of OA and presence of concomitant diseases. The analytic exercises for improvement of muscle strength in hip OA aim to influence the abductors, adductors, flexors and extensors of the hip joint. In knee OA, the target muscle groups are thigh muscles (quadriceps muscle and posterior group of thigh muscles), calf muscles (triceps surae muscle) as well as muscles around the hip joint. There are no established programmes for therapeutic exercises regarding the intensity, frequency, the degree and the interval for increasing of joint load. No significant difference in the efficacy of group vs. individual exercise programmes has been observed. In OARSI recommendations (2010), it is underlined that aerobic exercises and those for muscle strength have a moderate efficacy in knee OA for pain relief e.g. ES (effect size) - 0.52 (for aerobic exercises) and 0.32 (for strengthening exercises). The ES for functional improvement is 0.46 for the aerobic exercises and 0.32 for the strengthening exercises. The data for the role of therapeutic exercises in hip OA are scarce. An effect on pain has been observed (ES-0.38) but without improvement of functional capacity. Therapeutic exercises have low to moderate complex efficacy in patients with OA. They should be an obligatory part in the therapeutic regimens of the patients, which requires close collaboration between rheumatologists, general practitioners and physiotherapists as well as establishment of system for patient education.