The aim of this pilot study was to test the effectiveness and feasibility of a new, guided self-help program in five selected primary care patients with panic disorder (PD) or generalized anxiety disorder (GAD). Most patients achieved a clinically relevant improvement after 12 weeks of treatment with large effect sizes (Cohen's d) on all measures. Two patients could be considered as recovered according to the criteria of Jacobson and Truax (1991). Although all of the patients appeared to improve on all outcome measures, PD was easier to treat than GAD. These results suggest that a guided self-help manual treatment is an effective treatment possibility for PD and GAD in primary care. When these results can be replicated in a controlled trial in a larger sample, guided self-help treatment may be a first-line treatment in general practice in a step-by-step care approach of PD and GAD.