Background: Septic arthritis remains a challenging diagnosis in which the doctor often relies on laboratory tests. Objective: To examine the diagnostic utility of three ancillary tests-namely, white blood cells (WBC), erythrocyte sedimentation rate (ESR) and the WBC in the joint fluid (jWBC)-using likelihood ratios (LRs) and receiver operating characteristic (ROC) curves. Methods: This was a retrospective cohort study at the Jacobi Medical Center. Medical charts of patients who had undergone arthrocentesis were included. Patients who had ''dry taps'' were excluded from the study. Patients were considered to have septic arthritis if they had a positive arthrocentesis culture or operative findings. The primary outcomes of this study were the sensitivities, specificities, LR(+) and LR(2) values of the laboratory tests for septic arthritis. The performance characteristics of the laboratory tests were analysed using ROC curves. Results: 156 patients were enrolled, 16 (10%) had septic arthritis. The sensitivities for WBC, ESR and jWBC were 0.75, 0.75 and 0.50, and the specificities were 0.55, 0.11 and 0.88, respectively. The LR(+) values were 1.7, 0.84 and 4.0, and the LR(2) values were 0.46, 2.4 and 0.57, respectively. In ROC curve analysis, jWBC was the best test (area under the curve (AUC) 0.75, 95% confidence interval (CI) 0.58 to 0.92), followed by WBC (AUC 0.69, 95% CI 0.57 to 0.80) and ESR (AUC 0.55, 95% CI 0.37 to 0.74). A cut-off of jWBC = 17 500 maximised sensitivity and specificity on the ROC curve. Conclusions: jWBC was the best diagnostic test for septic arthritis, WBC and ESR were poor tests. However, no test was diagnostic, and the clinician must be careful with patients with a potential septic joint.