OBJECTIVE. The purpose of this work was to assess the annual incidence of arthritis in children and describe early disease and patient characteristics, microbiologic features, and immunogenetic factors in children with different subgroups of childhood arthritis. PATIENTS AND METHODS. A population-based multicenter study was performed in southeastern Norway between June 1, 2004, and May 31, 2005. The total population of children under 16 years of age was 255303. Physicians were asked to refer their patients with suspected arthritis to the local department of pediatrics or rheumatology. The children were assessed on the basis of clinical, radiologic, and laboratory examinations at inclusion and followed up at 6 weeks, 6 months, and thereafter as long as clinically indicated. A chart review was performed to identify patients with arthritis who had not been included prospectively. RESULTS. The total annual incidence of arthritis was 71 per 100000 children. Transient arthritis, juvenile idiopathic arthritis, postinfectious arthritis, and infectious arthritis were found in 43, 14, 9, and 5 of 100000 children, respectively. The incidence was higher in children under the age of 8 years than in older children (107 vs 34 per 100000). Arthritis occurred more frequently in boys than in girls before the age of 8 years but not thereafter. The median age of onset was lower in children with infectious arthritis than in those with other types of arthritis. Monarthritis was less frequent in patients with juvenile idiopathic arthritis than in the other subgroups (64% vs 83%-100%). Ten percent of the patients had poststreptococcal reactive arthritis, and only 1 had enteropathic arthritis. Autoantibodies and the presence of HLA-B27 were associated with juvenile idiopathic arthritis. CONCLUSIONS. The annual incidence of childhood arthritis was 71 per 100000 children. We found several factors that may help in differentiating between subgroups of arthritis.