The present study evaluated the initial effects of two types of parado
xical directives. Socially anxious males were randomly assigned to thr
ee groups: symptom prescription without reframing, symptom prescriptio
n with reframing, and a wait-list control condition. The assessment of
outcome consisted of various measures administered pretreatment, post
treatment, and at 1-month followup, as well as daily self-ratings that
were obtained throughout the course of the study. Subjects who were e
xposed to symptom prescription with reframing showed significantly gre
ater improvement than the subjects in the other groups, both at posttr
eatment and at followup. In contrast, the subjects who were exposed to
symptom prescription alone manifested only limited improvement and we
re not significantly less socially anxious after treatment than the su
bjects in the wait-list control group. The findings support the view t
hat paradoxical treatments are clinically useful when their effects in
clude changes in the meaning clients attribute to their symptoms.