To determine whether juvenile onset of spondylarthropathy is associated with specific features, a prospective, cross-sectional study comparing juvenile-onset and adult-onset spondylarthropathies was conducted in the Maghreb in 523 patients meeting Amor's criteria or the ESSG's criteria for spondylarthropathy. Demographic data and clinical findings at the time of inclusion and during the first two years of the disease were compared in the 437 patients with onset at 16 years of age or older and in the 86 patients with onset before 16 years of age using a Student's t test or a chi-square test. The risk of hip involvement during the course of the disease was estimated using Kaplan-Meier curves and compared in the two groups using a Cox model. Early in the disease, patients in the juvenile onset group were more likely to have peripheral arthritis (52% vs 39%, p = 0.021) and enthesopathies (55% vs 40%, p = 0.002) and less likely to have axial manifestations (41% vs 62%, p = 0.0001), as compared with the adult-onset group. These differences persisted after a follow-up of 9.2 years. Juvenile-onset disease was associated with a greater likelihood of hip involvement (54 +/- 6% vs 34 +/- 3% after ten years, p = 0.012). The male bias was more marked in the juvenile onset group (85%) than in the adult-onset group (72%) (p = 0.016). These data confirm that demographic characteristics, clinical manifestations, and disease severity differ between juvenile-onset and adult-onset spondylarthropathies.