Tennis elbow is due to a torque injury or sudden overstretching of tendons which insert into the epicondyles of the humerus. The predominant lesion is an enthesopathy--a pathological lesion at the insertion of tendon into bone. The most common site is at the lateral epicondyle and this is 3 times as frequent as at the medial epicondyle. Approximately 50% of tennis players can expect to get a tennis elbow at some time during their playing lifetime. In one-third of the players this will be severe enough to interfere with their tasks of daily living. The major unresolved question about the aetiology of tennis elbow is why it has its peak incidence between the ages of 40 and 50 years and why 90% of players then have no further recurrence. Making sense of the literature on the treatment of tennis elbow is difficult because there are few studies that have used the acceptable epidemiological techniques of the prospective randomised controlled trial or case-controlled study. Most papers are based on a collection of highly selected cases which represent the more intractable end of the tennis elbow spectrum and their reported results have been inconsistent. Tennis elbow is largely a self-limiting condition. The prime aim of treatment should be based on Hippocrates' first tenet of medicine--first do no harm. Therapy should start with the simple and conservative before progressing to the more complex and invasive therapies. It should be acceptable to the patient, cost-effective and where invasive therapy is recommended, the potential benefits should clearly outweigh the risks. The principles of therapy for tennis elbow are to relieve pain, microbleeding and inflammation, promote healing, rehabilitate the injured arm and try to prevent recurrence. The most effective modalities of treatment are found to be cryotherapy in the acute stage then nonsteroidal anti-inflammatory drugs and heat in its various modalities including ultrasound. This is combined with rest which is best defined as the absence of painful activity. Injection of a depot preparation of cortisone is effective although patient reports are not as flattering as those of doctors. There is no advantage and in fact considerable disadvantage in using more than 2 such injections. Therapies such as acupuncture and chiropractic have not been evaluated. Nevertheless they cause no harm, may result in good and should be tried before resorting to more invasive therapy. Rehabilitation should run parallel to treatment.(ABSTRACT TRUNCATED AT 400 WORDS)