Fluorescent antinuclear antibody (ANA) testing was performed on 141 sera from 114 patients with well defined rheumatic diseases including fibrositis syndrome and 24 sera from 24 healthy subjects using HEp-2 cells and rat liver as substrates. ANA titers were almost always higher on HEp-2, in most cases by 1:5 dilutions. ANA positivity or negativity was usually substrate independent, but there were exceptions. Two patients with SLE were ANA positive on HEp-2 only and rat liver only, respectively; patterns were homogeneous. Thirteen of 15 CREST patients had anticentromere antibodies, detected on HEp-2 only. "False-positive" ANA were invariably low titer, speckled and confined to one substrate.