Video-assisted thoracic surgery has been adopted by some thoracic surgeons as the preferred approach over thoracotomy for many benign and malignant diseases of the chest. However, little concrete evidence exists to support this technique as the superior approach. This randomized study was carried out to define the advantages of video-assisted lobectomy over muscle-sparing thoracotomy and lobectomy. Sixty-one patients with presumed clinical stage I non-small-cell lung cancer were entered into the study. Each patient was randomized to muscle-sparing thoracotomy and lobectomy or video-assisted lobectomy. Six patients were excluded from the study either because final pathologic results revealed nonmalignant disease (3 patients) or because an attempted video-assisted lobectomy was converted to a thoracotomy. This left 30 patients in the thoracotomy group and 25 patients in the video-assisted group. No significant differences existed between the two groups in operating time, intraoperative blood loss, duration of chest tube drainage, or length of hospital stay. Significantly more postoperative complications occurred in the thoracotomy group (p < 0.5), the majority of which were prolonged air leaks. Return to work time was not an issue because the majority of the patients were either retired or not working at the time of the operation. Only three patients had persistent postthoracotomy pain (thoracotomy, n = 2; video-assisted lobectomy, n = 1). We conclude that video-assisted lobectomy was not associated with a significant decrease in duration of chest tube drainage, length of hospital stay, postthoracotomy pain, or, in this group of patients, a faster recovery time and return to work. Video-assisted lobectomy continues to expose the patient to the risk of a major pulmonary resection being done in an essentially closed chest. These results illustrate the need for critical evaluation of video-assisted thoracic surgery before the procedure is accepted as a superior approach based on presumed and thus far unproved advantages.