Background: Psychometric properties of three functional assessment rating scales were compared for three types of target behaviours [self-injurious behaviour (SIB), stereotypic behaviour and aggressive/destructive behaviour]. Materials and method: The "Questions about Behavioural Function" (QABF), the "Functional Assessment for Multiple Causality" (FACT) and the "Functional Analysis Screening Tool" (FAST) were administered twice by two raters to 130 adults with intellectual disability (total of 29 raters). Results and conclusions: The reliability of the FACT and the QABF for estimates across all three behaviours was acceptable to good. Mean inter-rater reliability intra-class correlations across two administrations ranged from 0.63 to 0.68 for the QABF and from 0.65 to 0.78 for the FACT. Mean test-retest reliability for the QABF ranged from 0.81 to 0.82 and for the FACT from 0.86 to 0.87. Internal consistency across the subscales ranged from 0.89 to 0.96 for the QABF and from 0.92 to 0.96 for the FACT. The FAST had generally poorer reliability scores. Convergent and discriminant validity (Spearman [rho]) were better between FACT and the QABF than between the FAST and the other two instruments. (Contains 7 tables.)