Background: This study examined whether treatment response to stepped-care cognitive-behavioural treatment (CBT) is associated with changes in threat-related selective attention and its specific components in a large clinical sample of anxiety-disordered children. Methods: Ninety-one children with an anxiety disorder were included in the present study. Children received a standardized stepped-care CBT. Three treatment response groups were distinguished: initial responders (anxiety disorder free after phase one: child-focused CBT), secondary responders (anxiety disorder free after phase two: child-parent-focused CBT), and treatment non-responders. Treatment response was determined using a semi-structured clinical interview. Children performed a pictorial dot-probe task before and after stepped-care CBT (i.e., before phase one and after phase two CBT). Results: Changes in selective attention to severely threatening pictures, but not to mildly threatening pictures, were significantly associated with treatment success. At pre-treatment assessment, initial responders selectively attended away from severely threatening pictures, whereas secondary responders selectively attended toward severely threatening pictures. After stepped-care CBT, initial and secondary responders did not show any selectivity in the attentional processing of severely threatening pictures. Treatment non-responders did not show any changes in selective attention due to CBT. Conclusions: Initial and secondary treatment responders showed a reduction of their predisposition to selectively attend away or toward severely threatening pictures, respectively. Treatment non-responders did not show any changes in selective attention. The pictorial dot-probe task can be considered a potentially valuable tool in assigning children to appropriate treatment formats as well as for monitoring changes in selective attention during the course of CBT.