We studied the influence of demographic, clinical, laboratory and genetic features and radiographic damage at onset on the outcome after 2 years in a prospective study of 147 patients with classical or definite RA with disease duration shorter than 1 year at entry. Outcome was determined by physical disability and by radiographic damage of hands and feet. By means of multiple regression analysis and discriminant analysis outcome was explained from variables at the start and during the first 6 months. No clinically relevant conclusions could be made for physical disability due to the low explained variance and small number of patients with bad physical disability. Radiographic damage after 2 years was predicted by high disease activity at the start (measured as erythrocyte sedimentation rate, C-reactive protein or Disease Activity Score) combined with DR4 or DR2 (as a prognostically favourable factor) and rheumatoid factor positivity. Radiographic damage could be better predicted if disease activity during the first 6 months was included. Absence or presence of progression of radiographic damage could be correctly predicted in 83% of the patients.