Objective: The objective of this study was to discover autoantibodies to non-modified proteins associated with the presence/absence of ACPAs in RA. Methods: The autoantibody repertoire of 80 ACPA-negative and 80 ACPA-positive RA subjects from the Swedish population-based Epidemiological Investigation of RA (EIRA) cohort was screened using a suspension bead array built on protein fragments earlier described as autoimmunity targets. Four autoantibodies positive in the initial screening were validated in another set of EIRA samples containing 317 ACPA-positive, 302 ACPA-negative and 372 age- and sex-matched controls. The relationship between the four autoantibodies and lung abnormalities on high-resolution CT (HRCT) was examined in 93 early-RA patients from the LURA cohort. Association between the autoantibodies, smoking and MHC class II alleles was assessed by logistic regression analysis. Results: Anti-ANOS1 and anti-MURC IgG levels were associated with ACPA-positive status [odds ratio (OR) = 3.02; 95% CI 1.87-4.89; and OR = 1.86; 95% CI 1.16-2.97, respectively] and increased in ACPA-positive patients compared with controls. Anti-ANOS1 IgG was associated with smoking habit (OR = 2.11; 95% CI 1.22-3.69) and anti-MURC IgG with the presence of the MHC class II 'shared-epitope' genes (OR = 1.95; 95% CI 1.11-3.46). Anti-TSPYL4 IgG was associated with being ACPA negative (OR = 0.41; 95% CI 0.19-0.89). Anti-TSPYL4 IgG and anti-MAP2K6 IgG levels were increased in the ACPA-negative patients compared with controls. Presence of anti-MAP2K6 IgG and anti-TSPYL4 IgG correlated negatively with HRCT-defined lung abnormalities. Conclusion: These four autoantibodies may be useful in diagnostics and in predicting clinical phenotypes of RA.