Tests administered with accommodations to persons with disabilities have been considered nonequivalent to tests administered under standardized conditions to nondisabled test takers. This study examined the score change patterns for learning disabled students completing extended-time administrations of the SAT I: Reasoning Test in comparison to nondisabled students retesting under standard-time administrations. Results illustrate that learning disabled students generally performed about 0.5 of a standard deviation below nondisabled test takers. However, the mean score gain for learning disabled students first completing a standard-time SAT and retesting under an extended-time SAT was more than three times as large as the mean score gain for both nondisabled students testing under standardized conditions and learning disabled students testing with extended time on both occasions. An appendix contains supplementary tables. [This research was funded through the Joint Staff Research and Development Committee of ETS and the College Board.]