Diagnosing shoulder pain conditions is a challenging area of musculoskeletal practice. Subacromial impingement syndrome (SIS) is a clinical syndrome that indicates pain and pathology involving the subacromial bursa and rotator cuff tendons within the subacromial space. The three stages of SIS are subacromial bursitis, partial thickness and full thickness rotator cuff tears. The cause of SIS is believed to be multi-factorial with both extrinsic and intrinsic factors involved in its pathogenesis. Clinicians have traditionally diagnosed SIS using a clinical examination including a subjective history followed by confirmatory clinical tests. A review of the evidence for diagnostic accuracy of clinical tests highlights that individual tests have poor diagnostic accuracy. A combination of clinical tests or a clinical examination per se may be useful at ruling out rotator cuff tears, but is less accurate at detecting rotator cuff tears when it is present. There is consensus in the literature that particular combinations of signs and clinical features may be useful in diagnosing rotator cuff tears but not for diagnosing SIS. The vast majority of research to date examining the clinical diagnosis of SIS has been focused on individual clinical tests carried out by medical practitioners in specialist and tertiary care settings. This review has established that the majority of diagnostic accuracy studies for SIS and rotator cuff tears have had poor methodological design. This exploratory study was conducted with subjects undergoing a standardized clinical examination (index test) by a physiotherapist. The decision as to which specific tests were chosen for this research was based on supporting research within the literature and the test’s actual use within the New Zealand clinical setting. This included subjective history questions, active and passive shoulder movement tests and eleven SIS tests. Subjects were referred for a diagnostic ultrasound scan immediately following the clinical examination and results from the scan stood as the criterion reference standard. Thirty eight individuals (males n=23, females n=15) with new onset shoulder pain, who met the inclusion criteria, were assessed by a participating physiotherapist. Sensitivity, specificity, positive likelihood ratios, negative likelihood ratios and respective 95% confidence intervals were calculated for all variables of the examination. Individual variables from the clinical examination were tested for their association with the diagnostic ultrasound scan reference criterion using Pearson Chi-Squared Exact test. Potential predictor variables were retained as potential predictors for use in the logistic regression analysis to determine the most accurate set of clinical examination variables for diagnosing SIS and the individual pathological stages of SIS. The results indicate that no historical, subjective or objective features from the clinical examination are accurate in diagnosing SIS or rotator cuff tears. The presence of night pain demonstrated a significant correlation (P<0.02) with the criterion reference standard for the presence of subacromial bursa fluid/bunching. Night pain and pain with overhead activity has a high sensitivity for subacromial bursa fluid/bunching being present. The absence of night pain and the absence of pain with overhead activity are two subjective phenomena from a clinical examination that are useful in ruling out subacromial bursa fluid/bunching being present. Night pain was also found to be the best predictor of subacromial bursa fluid/bunching being present (P<0.012). Male gender (P<0.034) was the best predictor of partial thickness rotator cuff tears while being 60 years of age or older (P<0.01) significantly correlated with full thickness rotator cuff tears. The Drop Arm Sign (P<0.01) and External Rotation Lag Sign (P<0.01) were significantly correlated with SIS and full thickness rotator cuff tears. Clinical tests for all three pathological stages of SIS and subacromial bursa fluid/bunching being present, had equivalent or if not greater diagnostic accuracy than previous report studies in the literature. The Hawkins-Kennedy Test and Neer Sign can be used in the primary care setting to rule out the presence of subacromial bursa fluid/bunching or SIS if the tests are negative. For mid to end stage SIS (rotator cuff tears) the Empty Can Test and Drop Arm Sign with their high sensitivity can be used to rule out rotator cuff tears especially to the supraspinatus tendon when the tests are negative. Despite the small sample size and other limitations of this study, the findings are an important addition to the current literature surrounding the diagnostic accuracy of clinical tests for SIS and rotator cuff tears. This is the first study to use physiotherapists as examiners and to be set in a primary care setting. The study is also the first to examine the diagnostic accuracy of a range of historical and subjective features from the clinical examination. The results found in the current study could be used by future studies as a starting point in the development of a clinical decision or prediction rule to assist clinicians in the diagnosis of SIS and rotator cuff tears.