A survey of 24 speech-language pathologists was conducted to investiga
te the reliability of rating expressive language parameters in aphasia
. Ratings of the expressive language dimensions from the patient profi
le of the BDAE were made from spontaneous speech and sentence repetiti
on samples and were compared to fluency judgements for 10 different ap
hasic subjects. Agreement on a fluent/non-fluent diagnosis reached a c
riterion of two-thirds for only half of the subjects, despite the repo
rts of most clinicians that they used fluency classifications almost a
ll the time. A wide range of terms were used to describe each patient'
s language deficits. The distributions of ratings were also highly var
iable for individual subjects, especially on the dimensions of articul
ation and paraphasia rating. The results are explored for underlying c
ontributors to the variability of ratings observed. Implications for c
linical practice and research studies are discussed.