Objective <p>To assess the diagnostic accuracy of ultrasonography for meniscal injuries using arthroscopy as the reference standard.</p> Materials and methods <p>PubMed, Web of Science, and Cochrane Library databases were systematically explored for studies published since inception to July 2025. Studies using arthroscopy as the reference standard to assess the diagnostic accuracy of ultrasonography for meniscal tears met the inclusion criteria. Diagnostic accuracy, sensitivity, specificity, negative predictive value, and positive predictive value were extracted. Mean differences with 95% CI were calculated and the Inverse Variance fixed-effects model was used for meta-analysis.</p> Results <p>Eleven studies comprising 866 knees were included. Ultrasound demonstrated a high median diagnostic accuracy for both lateral (97%) and medial (94.2%) meniscal tears, with a pooled median accuracy of 95.3%. The greatest median sensitivity was for medial tears (93.1%), while lateral tears showed the highest specificity (95.6%). Radiologist-performed ultrasound showed superior diagnostic performance compared to surgeons, with a higher median sensitivity (90.9% vs. 85%) and specificity (93.9% vs. 85%). Meta-analysis of five studies comparing ultrasound and MRI revealed no significant difference in sensitivity or specificity for diagnosing medial (<i>P</i> = 0.81–0.92) or lateral (<i>P</i> = 0.86–0.91) meniscal tears. Heterogeneity was low-to-moderate for medial tears (<i>I</i>² = 0–52%; <i>P</i> = 0.08–0.82) and low for lateral tears (<i>I</i>² = 0%; <i>P</i> = 0.39–0.66).</p> Conclusion <p>When performed by experienced operators, ultrasound demonstrates high sensitivity and specificity for meniscal tears. However, these findings should be interpreted with caution due to limited information on tear location, morphology, and variability in study quality.</p>