Introduction - Because of the limited information on cervical spondylotic radiculopathy, we conducted a door-to-door two-phase survey in a Sicilian municipality. Material and methods - We first screened for cervical spondylotic radiculopathy among the inhabitants of the municipality: (N=7653, as of the prevalence day, November 1, 1987). Study neurologists then investigated those subjects suspected to have had a cervical spondylotic radiculopathy. Diagnoses were based on specified criteria. Results - We found 27 subjects affected by CSR (17 definite, 10 possible). Prevalence (cases per 1000 population) was 3.5 in the total population; it increased to a peak at age 50-59 years and decreased thereafter. The age-specific prevalence was consistently higher in women. Conclusions Comparison with other prevalence studies shows similar age-specific patterns, but different magnitudes, which may partly reflect methodologic differences across studies. (C) Munksgaard 1996.