This paper introduces a new method to identify output growth accelerations that integrates elements of both the “criteria-based” and “break-testing” approaches, which are prevalent in the literature. The proposed criteria do not impose a fixed length on growth accelerations, thus enabling duration analyses without relying on questionable statistical techniques for the identification of these accelerations. The findings show that growth accelerations last an average 13.4 years, albeit with significant variations in duration across regions. Initial conditions and contemporaneous domestic and external economic conditions all matter for the continuation of an acceleration, and changes in any single policy condition have less of an impact.