Previous research has established Internet-based cognitive behavioural therapy (CBT) for panic disorder (PD) as effective in reducing panic severity and frequency. There is evidence, however, that such programs are less effective at improving overall end-state functioning, defined by a PD clinician severity rating of [less than or equal to] 2 and panic free. In order to test the effect on end-state functioning of the incorporation of stress management material within a CBT program for PD, 32 people with PD were randomised to either Internet-based CBT (PO1), Internet-based CBT plus stress management (PO2) or an Internet-based information-only control condition (IC). Both CBT treatments were more effective at posttreatment assessment than the control condition in reducing PD severity, panic and agoraphobia-related cognition, negative affect and self-ratings of health. PO2 was more effective than PO1 at posttreatment assessment on PD severity and general anxiety, although at 3-month follow-up these differences were no longer apparent. This study provides further support for the efficacy of Internet-based CBT for PD and suggests that although the incorporation of stress management material confers short-term advantages over a standard program, it is not associated with any longer term improvements on panic severity and related cognitions, negative affect, general wellbeing and end-state functioning. (Contains 7 tables and 2 figures.)