The diagnosis of inflammatory joint disease rests on a constellation of symptoms, signs, laboratory test results and, occasionally, histological findings. Classification criteria have been developed by national learned societies, international panels of experts or, more rarely, an expert working alone. These criteria are intended to provide a common language for therapeutic trials and international publications. Yet, they are often inappropriately used as diagnostic tools for the individual patient. Identification of an early seroimmunologic marker with high sensitivity and specificity for classifying patients with recent-onset joint disease is a daunting challenge. Test performance characteristics such as sensitivity, specificity, positive and negative predictive values, and the positive or negative likelihood ratio help to assess the diagnostic usefulness of a laboratory test in a specific situation. The difference between the pretest and posttest likelihoods of obtaining a positive or negative result measures the usefulness, or performance, of a laboratory test in a specific situation according to the prevalence of the disease. A higher positive likelihood ratio indicates a more useful test. In a patient with inflammatory joint disease, the diagnosis can be sought by assaying a limited number of autoantibodies according to a decision tree. Thus, IgM rheumatoid factors (latex test or ELISA) and antibodies to filaggrin or other citrullinated proteins (antikeratin antibodies by indirect immunofluorescent assay or anticyclic citrullinated peptides by ELISA) identify more than 70% of cases of early rheumatoid arthritis with greater than 98% specificity. If these markers are negative, testing for antinuclear antibodies by indirect immunofluorescent assay on HEp-2 cells identifies 99% of cases of lupus and progressive systemic sclerosis. Confirmation of the diagnosis can be obtained by characterizing the autoantibodies: thus, presence of antidouble-stranded DNA (dsDNA, by the Farr radioimmunoassay, indirect immunofluorescent assay on Crithidia luciliae, or ELISA (IgG)) or of antinucleosome antibodies (ELISA) indicates lupus, whereas anticentromere, antitopoisomerase I (Scl 70), and antinucleolar antibodies point to progressive systemic sclerosis. A positive test for antibodies to soluble nuclear antigens of the U1 RNP type suggests mixed connective tissue disease or lupus but may indicate scleroderma. Anti-Sm antibodies are found in fewer than 10% of lupus patients but are highly specific. Anti-SSA (Ro) and anti-SSB (La) suggest lupus or primary Sjögren's syndrome. When tests are negative for ANA, several antibodies to cytoplasmic organelles are valuable diagnostic tools, such as anti-J01 for polymyositis syndromes and antiribosome antibodies for lupus, although their sensitivity is modest (20-25%). Finally antineutrophil cytoplasmic antibodies (ANCAs) ensure the diagnosis of small-vessel vasculitides, which often involve the lungs and kidneys. Thus, in diffuse Wegener's granulomatosis, ANCAs exhibiting the classic cytoplasmic pattern and corresponding by ELISA to anti-PR3 are found. In microscopic polyangiitis the ANCAs are peripheral and correspond by ELISA to antimyeloperoxidase antibodies. Tests for other antibodies are less often needed to evaluate inflammatory joint disease.