In a prospective comparison, 60 patients suffering from complicated spontaneous pneumothorax were randomly allocated to receive treatment by a video-assisted thoracoscopic surgery (VATS) technique or by thoracotomy. Thirty patients underwent bullectomy and apical pleurectomy by VATS performed through three 2-cm incisions (group V) and 30 patients underwent a similar surgical procedure through a posterolateral thoracotomy (group T). The median operating time was significantly longer in group V (45 versus 37.5 minutes; p < 0.05), but the postoperative analgesic requirement and hospital stay were less than those in group T. On the third postoperative day, the reductions in the forced expiratory volume in 1 second and forced vital capacity were significantly lower in group V than in group T (p < 0.05 and p < 0.01, respectively). Initial treatment of the spontaneous pneumothorax was effective in 27 patients (90%) in group V and in 29 patients (97%) in group T. There have been two late recurrences in group V and one in group T at a median follow-up of 15.1 months and 16.3 months, respectively. Within the study group, 30 consecutive patients presented with primary spontaneous pneumothorax. In this subgroup there was no significant difference in the operating time between VATS and thoracotomy, but postoperative pain, hospital stay, and pulmonary dysfunction were all less for those undergoing VATS. All treatment failures were in the subgroup of 30 consecutive patients who presented with secondary spontaneous pneumothorax, and the hospital stay in this group was prolonged by the use of VATS. We conclude from our findings that VATS is superior to thoracotomy in the treatment of primary spontaneous pneumothorax.(ABSTRACT TRUNCATED AT 250 WORDS)