Prior systematic reviews of rehabilitation for nondescript shoulder pain have not yielded clinically applicable results for those patients with subacromial impingement syndrome (SAIS). The purpose of this study was to examine the evidence for rehabilitation interventions for SAIS. The authors used data source as the method. The computerized bibliographic databases of Medline, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), and the Cochrane Database of Systematic Reviews were searched from 1966 up to and including October 2003. Key words used were "shoulder," "shoulder impingement syndrome," "bursitis," and "rotator cuff" combined with "rehabilitation," "physical therapy," "electrotherapy," "ultrasound," "acupuncture," and "exercise," limited to clinical trials. Randomized clinical trials that investigated physical interventions used in the rehabilitation of patients with SAIS with clinically relevant outcome measures of pain and quality of life were selected. The search resulted in 635 potential studies, 12 meeting inclusion criteria. Two independent reviewers graded all 12 trials with a quality checklist averaged for a final quality score. The mean quality score for 12 trials was 37.6 out of a possible 69 points. Various treatments were evaluated: exercise in six trials, joint mobilizations in two trials, laser in three trials, ultrasound in two trials, and acupuncture in two trials. The limited evidence currently available suggests that exercise and joint mobilizations are efficacious for patients with SAIS. Laser therapy appears to be of benefit only when used in isolation, not in combination with therapeutic exercise. Ultrasound is of no benefit, and acupuncture trials present equivocal evidence. The low to mediocre methodologic quality, small sample sizes, and general lack of long-term follow-up limit these findings for the development of useful clinical practice guidelines. Further trials are needed to investigate these rehabilitation interventions, the superiority of one intervention over another, and the long-term outcomes of rehabilitation. Moreover, it is imperative that clinical guidelines are developed to indicate those patients who are likely to respond to rehabilitation.