BACKGROUND: Ultrasound (US) guidance is widely used for needle positioning for cervical medial branch blocks (CMBB) and radiofrequency ablation, however, limited research is available comparing different approaches. OBJECTIVE: We aimed to assess the accuracy and safety of 3 different US-guided approaches for CMBB. STUDY DESIGN: A cadaveric study divided into ultrasound-guided needle placement and fluoroscopy evaluation stages. SETTING: Department of Pathology, Forensic, and Insurance Medicine, Semmelweis University. METHODS: Sonographically guided third occipital nerve (TON), C3, C4, C5 and C6 medial branch injections and radiology evaluations were performed. The 3 approaches compared were: 1. ES (published by Eichenberger-Siegenthaler): US probe in the coronal plane to visualize the cervical articular pillars, needle approach out of the plane, from anterior to posterior. 2. Fi (published by Finlayson): US probe in the transverse plane to visualize a cervical articular pillar and its lamina, needle approach in the plane, from posterior to anterior. 3. FiM (Modified Finlayson approach): Needles are placed as in Fi, but then adjusted with a coronal view of the cervical articular pillars. Fluoroscopy images were taken and later evaluated, for “crude”, “high precision” and “dangerous” placement. RESULTS: One hundred and fifty-five needle placements were assessed (10 were excluded, as no anterior-posterior fluoroscopy images were saved). Interobserver agreement on position of needle placement between the 5 observers was very high; the Fleiss’ Kappa was 0.921. For crude placement, no significant differences were identified between various approaches; (77.6%, 79.5%, and 75.6% for the ES, Fi, and FiM respectively). However, for placement in predefined high-precision zones, ES resulted in significantly more success (ES: 42.9%, Fi: 22.7%, and FiM: 24.4%, P = 0.032). Fi and FiM resulted in no dangerous placements, while ES led to the potential compromise of the exiting nerve root and vertebral artery on three occasions. In 10% of the placements, the levels were identified wrongly, with no difference between the various approaches. LIMITATIONS: Feedback from a live patient, may prevent some existing nerve root injections, unlike in a cadaver. Though a higher number of needles were placed in this study than in most available publications, the number is still low at each individual medial branch level. CONCLUSION: Fi proved safer than ES. Fi was equally successful in targeting the articular pillar, however, ES proved the most successful in placing the needle in the center of the articular pillar. Adding another, (coronal) US view to check needle position in FiM did not improve safety or precision. Identifying CMB levels with the US is challenging with all approaches, therefore we still recommend using fluoroscopy for level identification. While there were pros and cons with either procedure, the efficacy findings of previous papers were not replicated on elderly cadavers with arthritic necks. KEY WORDS: Cervical facet, medial branch block, safety, accuracy, ultrasound, fluoroscopy