The objective of the study was to assess the long-term efficacy of fluoroscopically-guided cervical nerve root block as a non-surgical treatment for cervical radicular pain. This was a retrospective study of 19 consecutive patients who had undergone cervical nerve root blocks over a period of 18 months, at a regional neurosurgery referral centre in the UK. Two of these patients underwent a second procedure; therefore, the number of total nerve root blocks was 21. Data regarding age, sex and diagnosis were obtained from medical records. MR reports formed the basis for imaging findings. Patients were contacted by telephone and post in order to obtain information about their 'pain relief. This was measured by using a 100-point Visual Analogue Scale (VAS). Four points in time were chosen in order to determine the time course of pain relief, i.e. before procedure, at 2 weeks, at 2 months and at 6 months following the procedure. Mean VAS scores at 6 month follow-up were broken up into 3 categories to indicate the level of pain relief. These categories were: VAS decrease of less than 20 points indicating no relief (12 procedures, 57.1%); VAS decrease 20 - 40 points, i.e. moderate relief (three procedures, 14.3%); VAS decrease of greater than 40 points, i.e. significant relief (six procedures, 28.6%). CNRB has limited efficacy for definitive treatment of nerve root pain, but may lead to significant short term relief, in a subgroup of such patients.