Object To compare therapeutic efficacy and safety of ultrasound (US)-guided selective nerve root block (SNRB) and fluoroscopy (FL)-guided transforaminal epidural steroid injection (TFESI) for cervical spine radiculopathy (CSR). Method 156 patients with CSR randomly received US-guided SNRB verified by FL or FL-guided TFESI. We hypothesised that the accuracy rate of contrast dispersion into epidural or intervertebral foraminal space in the US group was not inferior to that in the FL group with a margin of clinical unimportance of -15%. Pain intensity assessed by Numeric Rating Scales (NRS) and functional disability estimated by neck disability index (NDI) were compared before treatment, at 1, 3 and 6 months after the intervention. Puncture time and complication frequencies were also reported. Results 88.7% and 90.3% accuracy ratings were respectively achieved in the US and FL groups with a treatment difference of -1.6% (95%CI: -9.7%, 6.6%) revealing that the lower limit was above the non-inferiority margin. Both NRS and NDI scores illustrated improvements at 1, 3 and 6 months after intervention with no statistically significant differences between the two groups (all p > .05). Additionally, shorter administration duration was observed in the US group (p < .001). No severe complications were observed in both group. Conclusion Compared with the FL group, the US group provided a non-inferior accuracy rate of epidural/foraminal contrast pattern. For the treatment of CSR, the US technique provided similar pain relief and functional improvements while facilitating distinguishing critical vessels adjacent to the foramen and requiring a shorter procedure duration without exposure to radiation. Therefore, it was an attractive alternative to the conventional FL method. Key messages We conducted a prospective, open-label, randomised and non-inferiority clinical trial to estimate a hypothesis that the precisely accurate delivery through ultrasound (US)-guided cervical selective nerve root block (SNRB) was non-inferior to that using FL-guided transforaminal epidural steroid injection. Additionally, US-guided SNRB was as effective as FL-guided TFESI in the treatment effect on pain relief and function improvements. Notably, the US technique might be an alternative to the conventional FL method due to the ability to prevent inadvertent vascular puncture (VP) and intravascular injection (IVI) with a shorter administration time and absence of radiation exposure.