The updated international consensus criteria for definite antiphospholipid syndrome (APS) are useful for scientific clinical studies. However, there remains a need for diagnostic criteria for routine clinical use. We audited the results of routine antiphospholipid antibodies (aPLs) in a cohort of 193 consecutive patients with aPL positivity-based testing for lupus anticoagulant (LA), IgG and IgM anticardiolipin (aCL) and anti-ß(2)glycoprotein-1 antibodies (aß(2)GPI). Medium/high-titre aCL/aβ(2)GPI was defined as >99th percentile. Low-titre aCL/aβ(2)GPI positivity (>95(th )< 99(th) percentile) was considered positive for obstetric but not for thrombotic APS. One hundred of the 145 patients fulfilled both clinical and laboratory criteria for definite APS. Twenty-six women with purely obstetric APS had persistent low-titre aCL and/or aβ(2)GPI. With the inclusion of these patients, 126 of the 145 patients were considered to have APS. Sixty-seven out of 126 patients were LA-negative, of whom 12 had aCL only, 37 had aβ(2)GPI only and 18 positive were for both. The omission of aCL or aβ(2)GPI testing from investigation of APS would have led to a failure to diagnose APS in 9.5% and 29.4% of patients, respectively. Our data suggest that LA, aCL and aβ(2)GPI testing are all required for the accurate diagnosis of APS and that low-titre antibodies should be included in the diagnosis of obstetric APS.