Although there are reports that the State-Trait Anxiety Inventory (STAI) should be adapted to older adults, the standard version of the instrument is consistently used with this population. Bouchard, Gauthier, Ivers and Paradis (1996) have adapted a French version of the STAI for a population of older adults and found one item with extremely low item-remainder correlation. In Study 1 (N = 57), alternative formulations of item 24 were assessed to examine if the low item-remainder correlation was related to problems in translation that could become apparent in a sample of older adults. Study 2 (N= 188) was conducted in order to replicate the findings of Study 1 and assess the factor structure of the instrument. In Study 3, 46 older adults completed the instrument on two occasions with a 35-day interval to assess test-retest reliability. Our results suggest that: (a) item 24 should be removed from the trait anxiety scale and be replaced by the mean of the other anxiety-present items; (b) the instrument has a four-factor structure similar to what is found with the standard version of the STAI in nonelderly samples; and (c) both subscales are highly stable.