Compare the effectiveness of catheter-directed cervical interlaminar epidural steroid injection (C-CIESI) with triamcinolone to cervical transforaminal steroid injection (CTFESI) with dexamethasone for the treatment of refractory unilateral radicular pain. Prospective, randomized, comparative trial. Primary outcome: proportion of participants with ≥50% numeric rating scale pain score reduction from baseline "dominant pain" (the greater of arm vs neck) at one month postinjection. Secondary outcomes: ≥30% Neck Disability Index (NDI-5) reduction and Patient Global Impression of Change (PGIC) response indicating "much improved" or "very much improved." One hundred twenty participants (55.6% females, 52.3 ± 12.5 years of age, BMI 28.2 ± 6.5 kg/m2), were enrolled. The proportions of participants who experienced ≥50% pain reduction at one, three, and six months were 68.5% (95% CI = 54.9-79.5%), 59.3% (95% CI = 45.7-71.6%), and 60.8% (95% CI = 46.7-73.2%), respectively, in the C-CIESI group compared with 49.1% (95% CI = 36.4-62.0%), 46.4% (95% CI = 33.8-59.6%), and 51.9% (95% CI = 38.4-65.2%), respectively, in the CTFESI group. The between-group difference at one month was significant (P = 0.038). The proportions of participants who experienced a ≥30% NDI-5 score improvement were 64.0% (95% CI = 49.8-76.1%) and 54.9% (95% CI = 41.1-68.0%) in the C-CIESI and CTFESI groups (P = 0.352). Participants reported similar PGIC improvement in both groups: At six months, 53.2% (95% CI 38.9-67.1%) and 54.5% (95% CI = 39.7-68.7%) of the C-CIESI and CTFESI groups reported being "much improved" or "very much improved," respectively (P = 0.897). Both C-CIESI with triamcinolone and CTFESI with dexamethasone are effective in reducing pain and disability associated with refractory unilateral cervical radiculopathy in a substantial proportion of participants for at least six months.