This experimental study is to assess the effectiveness of structured group counselling, using a phenomenological-cognitive behavioural apptoach, on pupils who were bullies. Two types of group counselling were developed, 'the weekly group counselling' and 'the mararhon group counselling'. The effectiveness of group counselling in reducing bullying behaviour was measured using four self-report scales related to bullying behaviour, emphaty, aggressiveness and self-esteem. The subjects (n=48), comprising of Form 2 and Form 4 pupils, were randomly assigned to two Form 2 and two Form 4 experimental groups and two wait-list control groups. Subjects were administered the pre- and post-test for the four self-report measures. Treatment consisted of weekly group counselling and marathon group counselling. Altogether experiment group subjects received 11 counselling sessions of two-hours duration per session, over six weeks. The treatment focused on the developing new belief systems and building emphaty through cognitive restructuring and role-play. Statiscal analyses the experimental data were, Multivariate analyses of covariance-Mancova, and Post Hoc-Turkey. The significatn level was set at .05. The study showed that the cognitive behaviour group counselling based on the integrated model of phenomenological-cognitive bahavioural approach was significantly effective in: i) reducing the bullying behaviour and aggressiveness of all the subjects; ii) increasing the empathy of all subjects and iii) increasing the self-esteem of Form 2 subjects. However, it had no significant effect in increasing the self-esteem on Form 4 subjects. Group counselling can help reduce bullying while increasing the emphaty.