Purpose <p>Pathological diagnosis is important for the treatment of deep suprahyoid head and neck lesions, and tissue sampling needs to balance minimal invasiveness and accuracy. The purpose of this study was to evaluate diagnostic accuracy and factors associated with diagnostic failure of core needle biopsy (CNB) with CT-guided in deep suprahyoid head and neck lesions.</p> Methods <p>The records of 204 patients who underwent CT-guided CNB were retrospectively reviewed. CT-guided CNB was conducted for pathological diagnosis with the use of 18-G coaxial biopsy needles. Diagnostic accuracy for the diagnosis of lesions were calculated by comparing the biopsy results with the operative specimen or based on treatment response and clinical follow-up more than 6 months. Factors associated with biopsy failure was identified by chi-square test and logistics regression of procedure characteristics and lesion features.</p> Result <p>All 204 specimens were deemed adequate for histological diagnosis, with no immediate or delayed severe complications encountered. The diagnostic performance showed a sensitivity of 89.2% (141/158), specificity of 97.8% (45/46), and overall accuracy of 91.2% (186/204). Respectively, lesions with poorly defined margins or pre-procedural diagnostic imaging were the potential factor for diagnostic failure.</p> Conclusion <p>CNB with CT-guidance is an effective procedure for tissue diagnosis of patient with primary deep suprahyoid head and neck lesions and skull base lesions. Notably, lesions with poorly defined margins and suboptimal pre-procedural imaging emerged as potential factors contributing to diagnostic failure. Specifically, for lesions with indistinct boundaries-wherein the extent of the lesion is difficult to delineate—pre-procedural assessment using magnetic resonance imaging is recommended to enhance the clarity of lesion margins, thereby potentially improving diagnostic accuracy.</p>