A follow-up blind study, of the ability of "cross-reactive" antisera to distinguish between the cells of Dutch patients with ankylosing spondylitis (AS) and normal controls, was performed in Leiden. Of the 45 cell samples tested, 29 were fresh peripheral blood mononuclear (PBM) cells while 15 were cryopreserved PBM. No false positives but one false negative was identified among the 45 samples, and the "negative" was confirmed after the recoded cryopreserved cells from this patient were retested. It is concluded that the "cross-reactive" antisera raised in Sydney give good discrimination between patients and normals. Factors affecting the success of the 51Cr-release cytotoxicity assay, and possible reasons for the failure of others to confirm these observations, are briefly discussed.