We investigated the relationship between clinical, laboratory and genetic markers and outcome measures in 159 patients with recent onset of inflammatory arthritis (IA). The majority of patients were managed in community-based rheumatology practice. Median duration of arthritis at baseline was 3 months with median follow-up of 4.0 years (range 0-10). Markers of disease activity and 1987 ACR criteria for rheumatoid arthritis (RA) were estimated every 6 months for the first 2 years and annually thereafter. Presence of shared epitopes (SE) was established by PCR-based method. Main outcome variables were attainment of remission and presence of erosions on X-rays of hands and feet at 3 years. Remission was seen in 34.3% of patients and was independently related to age 60 and older (odds ratio (OR) 3.2; 95% confidence interval (CI), 1.2-8.7) and inversely to the presence of rheumatoid factor (RF) (OR 8.3; 95% CI, 3.2-21.3 for persistent arthritis). Patients with two SE were likely to have persistent arthritis (P=0.006), but this was not significant when corrected for RF. Independent predictors for erosions at 3 years were RF (OR 7.5; 95% CI, 1.9-29.5) and area under the curve for number of swollen joints (OR 1.08; 95% CI, 1.02-1.16). SE status was not predictive of erosions at 3 years (OR 1.6; 95% CI, 0.7-3.7). In univariate analysis, patients possessing DERAA motif on DRB1 were less likely to have erosive disease than without this motif at 4 years (OR 0.21; 95% CI, 0.0-0.9, P=0.037) but this finding was partly explained by adjusting for RF (adjusted OR 0.24; 95% CI 0.04-1.37). In this study of recent onset IA, active disease and RF were associated with poor outcome. Whilst SE did not predict erosive disease, patients with DERAA motif may be protected against erosions whilst the presence of two SE alleles suggests persistence of arthritis.