Crawford, Allan, Cochrane, and Parker (1990) used demographic variables in a multiple regression equation to predict NART error scores in a general population sample. The present study examined the applicability of Crawford's equation to neurological patients. Using this equation and the authors' suggestion that NART performance can be impaired in neurological patients, the data were also examined in participants with diffuse cortical damage. Participants (n = 175) were patients referred for clinical neuropsychological evaluation in north-east England. The findings confirmed the usefulness of Crawford's equation: Participants showing significantly higher obtained than predicted NART error scores (identified using Crawford's equation), were categorised as NART-impaired. Analyses of NART-impaired and NART-unimpaired patient subgroups showing diffuse cortical damage indicated the former had a significantly higher mean NART error score and mean discrepancy between NART-predicted and observed IQs. However, a significantly lower mean verbal (VIQ) IQ score was also noted in the NART-impaired participants, suggesting that the risk of an impaired NART performance is greater in a diffusely damaged patient with a lower premorbid VIQ.