To determine the impact of the HLA B27 antigen on the expression of spondylarthropathies, we conducted a retrospective cross-sectional study of the 116 spondylarthropathy patients whose HLA B27 phenotype was determined during a stay in the Morvan Hospital rheumatology department, Brest, France, between January 1, 1986, and December 31, 1994. Age at disease onset was younger in the HLA B27-positive patients (31.5 +/- 14 years versus 40 +/- 15 years; p = 0.008), who were more likely to have buttock pain (odds ratio, 4.84; p < 0.001) and roentgenographic evidence of sacroiliitis (odds ratio, 5.34; p < 0.001) and less likely to have psoriasis (odds ratio, 0.15; p < 0.0001), as compared with their HLA B27-negative counterparts. Peripheral arthritis occurred in similar proportions of patients with and without the HLA B27 antigen. Presence of HLA B27 was of little value for the diagnosis of spondylarthropathy in patients with inflammatory joint disease involving peripheral joints (sensitivity 50%, specificity 92%, positive predictive value 53%, negative predictive value 91%). Higher mean values of the erythrocyte sedimentation rate (40.3 versus 30.6 mm/h; p < 0.05) and serum C-reactive protein level (29.8 versus 16.8 mg/L, p < 0.005) were seen in patients with the HLA B27 antigen. Our data from patients with any form of spondylarthropathy show that the HLA B27 antigen is associated with earlier-onset disease, involvement of the sacroiliac joints and more severe inflammation.