Minimal-access surgery (MAS) is rapidly becoming the surgical approach of choice for a variety of surgical disorders in adults, but its use in children remains a relative novelty. Most pediatric surgeons continue to harbor justifiable concerns about the morbidity of this modality owing to the cumbersome nature of the instruments and the technical difficulty associated with two-dimensional views. The purpose of this study was to determine the complication rate end the lessons learned from the use of MAS in performing a variety of procedures in a large series of children. To determine complications, the authors reviewed the medical records of all children (n = 636; age range, 1 month to 19 years) who underwent laparoscopy (LAP) or thoracoscopy (THO) during a 5-year period (January 1, 1990 through December 31, 1994). The follow-up ranged from 1 week to 45 months. THO was performed in 62 children. Conversion to thoracotomy occurred in eight children (13%), because of inability to localize the lesion (3), unresectibility (2), inadequate tissue sample (1), unsafe access (1), hypoxemia (1), or inadvertent esophagotomy (1). Postoperatively, two ventilator-dependent children had tension pneumothorax after lung resection and required chest tubes. LAP was performed on 574 children, with conversion to laparotomy occurring in 15 (2.6%), because of technical reasons (10) or intraoperative complications (5). The complication rate of LAP was 2% (12 of 574). Early in the experience, intraoperative complications that led to laparotomy included hemorrhage during appendectomy (2), cholecystectomy (1), and splenectomy (1); and esophagotomy during a fundoplication (1). Other technical problems in the postoperative period were a malpositioned Nissen fundoplication and a gastric volvulus after gastrostomy and Nissen fundoplication owing to improper gastrostomy tube position. In addition, two children had a hernia at the umbilical trocar site that had been used for contralateral inguinal exploration, and cellulitis developed in three patients when a gastrostomy tube was brought out through a trocar site. Other complications not specific to MAS included pelvic abscess after appendectomy (5); small bowel obstruction after jejunostomy catheter placement (1) and combined cholecystectomy/appendectomy (1); enterocolitis (1) and severe hyponatremia (1) after pull-through for Hirschsprung's disease; and pneumonia after splenectomy (1). The overall complication rate of MAS was 4% (25 of 626), and there were no deaths. The initial use of MAS was associated with technical errors, which decreased with experience. Based on this study, the authors recommend (1) routine placement of a thoracostomy tube in children after THO if they require postoperative ventilator support; (2) using the open hernia sac to place a 70° telescope for contralateral inguinal exploration; and (3) not using a trocar site for gastrostomy tube placement in immune-suppressed patients. With appropriate training and experience, MAS can be used safely in children, for a wide variety of diseases, with minimal morbidity and mortality.