Pathologic extranodal extension (pENE) impacts treatment planning and is an important prognostic indicator for patients with head and neck squamous cell carcinoma (HNSCC). Computed tomography (CT) is a commonly used modality for assessment of radiographic ENE (rENE). To determine the predictive value of CT‐identified rENE in predicting pENE, we performed a systematic review through a search of 4 databases (PubMed, Scopus, Cochrane, and OVID). Meta‐analysis of diagnostic performance based on human papillomavirus (HPV) status was conducted. For HPV‐negative HNSCC, pooled sensitivity, specificity, and accuracy were 60.6%, 93.3%, and 82.6%, respectively. Overall positive predictive value (PPV) was 82.7%. For HPV‐positive HNSCC, pooled sensitivity, specificity, and accuracy were 77.7%, 72.2%, and 63.8%, respectively. Overall PPV was 68.6%. Significant differences were observed in diagnostic performance parameters between the two cohorts. The radiographic characteristics of HPV‐positive and HPV‐negative nodal metastases in HNSCC differ and radiographic evaluation of ENE in HPV‐positive nodes is challenging. Development of refined imaging characteristics of HPV‐positive nodes is needed to improve diagnostic performance.