Sera of patients with rheumatic diseases often contain autoantibodies to nuclear antigens. Problems of the specificity of autoantibodies and of their usage for studying nuclear components and of the whole nuclear organization are reviewed.