This study investigated the long-term follow-up effectiveness of (cognitive-) behavioural group and individual treatments for generalized social phobia. Patients were reassessed 18 months after they had finished one of the following treatment packages: (1) exposure in vivo; (2) cognitive therapy followed by exposure in vivo; or (3) a cognitive-behavioural treatment in which both strategies were integrated from the start. Half of the patients were individually treated, the other half in a group. Self-report assessments were held before and after treatment and at 3-month and 18-month follow-ups. Repeated measures MANOVAs on the patients who completed the long-term follow-up (n = 50) demonstrated significant time effects, indicating lasting improvement compared with the pretest. Between the posttest and the 18-month follow-up no significant changes were observed. ANCOVAs either with the pretest or the posttest as covariate showed a significant interaction at 18-month follow-up between treatment package and treatment modality on three of the four compound outcome variables. The group treatment with exposure in vivo alone had been the most effective in the longer term, the integrated group treatment the least effective, while the individual treatments had given improvements in-between. Results are qualified in view of numbers of dropouts, additional treatments in the respective treatment conditions, and clinical relevance.