The Salford Objective Recognition Test (SORT) was developed as a measu
re of recognition memory suitable for use with elderly people as it re
quires a shorter administration time than Warrington's Recognition Mem
ory Test. Normative data for the SORT were collected from 50 normal el
derly people living in the community. Fifty stroke patients were asses
sed on the SORT and other tests of psychological functioning. The word
s and faces subtests were found to correlate well with other measures
of memory. The SORT was found to be uninfluenced by the effects of sex
and time since onset of stroke, but was affected by language difficul
ties and perception problems. The split-half reliability of the test w
as found to be good as alternative versions of the subtests correlated
well with each other.