<b><i>Background:</i></b> Cognitive-behavioural therapy (CBT) is the first-choice treatment in clients with ultra-high risk (UHR) for psychosis. However, CBT is an umbrella term for a plethora of different strategies, and little is known about the association between the intensity and content of CBT and the severity of symptomatic outcome. <b><i>Methods:</i></b> A sample of 268 UHR participants received 6 months of CBT with case management (CBCM) in the context of the multi-centre NEURAPRO trial with monthly assessments of attenuated psychotic symptoms (APS). Using multilevel regressions and controlling for the initial severity of APS, the associations between (1) number of CBCM sessions received and severity of APS and (2) specific CBCM components and severity of APS were investigated. <b><i>Results:</i></b> In month 1, a higher number of sessions and more assessment of symptoms predicted an increase in APS, while in month 3, a higher number of sessions and more monitoring predicted a decrease in the level of APS. More therapeutic focus on APS predicted an overall increase in APS. <b><i>Conclusions:</i></b> Our findings indicate that the association between intensity/content of CBCM and severity of APS in a sample of UHR participants depends on the length of time in treatment. CBCM may positively impact the severity of APS later in the course of treatment. Therefore, it would seem important to keep UHR young people engaged in treatment beyond this initial period. Regarding the specific content of CBCM, a therapeutic focus on APS may not necessarily be beneficial in reducing the severity of APS, a possibility in need of further investigation.