This retrospective study was undertaken to determine the revision rate for dynamic sphincter pharyngoplasty (DSP) at the University of Michigan Medical Center to analyze the determinants contributing to the need for revision pharyngoplasties, and ultimately to improve primary pharyngoplasty to avoid the need for revision. The records of 30 children with repaired palatal clefts who presented with velopharyngeal insufficiency and hypernasal speech, and who underwent DSP from January 1988 through July 1994 were reviewed. Clinical follow-up ranged from 6 to 48 months (mean, 20.2 months). Seven of the original 30 patients (23%) had persistent, moderate-to-severe hypernasality that required reoperation, while 1 patient (3%) demonstrated hyponasality requiring revision. Seven of 8 patients who underwent revision pharyngoplasty had acceptable speech after revision. Dehiscences, low-lying pharyngoplasty flaps, and end-to-end suturing of the flaps were the main determinants resulting in the need for revision. In our study, female gender and older age was associated with a higher success of primary operation.