Research on L2 acquisition of the Spanish perfective and imperfective past has suggested that order-of-instruction (preterite before imperfect) may significantly contribute to learners' difficulty with mastering the imperfect. We sought to empirically test the effect of order-of-instruction by implementing a program-wide intervention in a beginner Spanish program at a large US university (N = 697 students) which reverted the timing of teaching the preterite and imperfect (imperfect before preterite). Except for the order in which the tenses were taught, all aspects of the curriculum remained identical across the "preterite-first" and the "imperfect-first" groups. Results from a cloze test and an analysis of student narrative texts demonstrate that teaching the imperfect first had neither a significant effect on suppliance nor on appropriate use of preterite-imperfect forms, thus rejecting the hypothesis that order-of-instruction plays a central role in L2 Spanish tense-aspect development.