To reassess the role and timing of operative intervention for spontaneous pneumothorax, 119 patients were retrospectively reviewed to compare recurrences, complications, and hospital stay between a nonoperative group (Group 1) and an operative group (Group 2). Total hospital days were greater in Group 2, but excluding the length of preoperative stay, the number of hospital days were similar in both groups. Group 1 patients more than 40 years old had a longer postoperative hospitalization, but not a higher rate of complication. Overall, morbidity was not different between the two groups, and there were no immediate or perioperative deaths in either group. There were no recurrences in Group 2. At least 11 of the 49 patients in Group 1 had a recurrence (p = .012). Considering the excellent results achieved with operative pleurodesis and the total hospital days accrued with nonoperative therapy, operative pleurodesis should be considered if an active leak persists more than three days after the initial episode of spontaneous pneumothorax or at the time of the first recurrence in the hospitalized patient.