The 'parent distortion' hypothesis regarding assessment of a child's anxiety state was examined in the mothers of 128 boys with an autism spectrum disorder. Mothers' own generalised anxiety disorder (GAD) and their ratings of their sons' GAD were compared with the boys' self-ratings and the ratings of the boys given by a clinician. Boys' cortisol concentrations were also explored for their association with these three sources of GAD ratings. Results indicated that mothers' GAD was significantly and directly correlated with the ratings they gave for their sons' but that only mothers who were above-minimally anxious gave ratings of their sons' anxiety that significantly agreed with those from the clinicians. Minimally-anxious mothers appeared to underestimate their sons' anxiety, and these effects generalised to their sons' self-ratings of anxiety. Associations between the boys' cortisol concentrations suggested an interaction between the presence of the diurnal fluctuation in boys' cortisol concentrations and mothers' anxiety states.