The present study extends previous research (e.g., Sanz, 1995; VanPatten & Cadiemo, 1993; VanPatten & Sanz, 1995) on the relative effects of two types of instruction (i.e., input processing and output processing) on the use of Spanish clitic pronouns. Thirty-three classroom learners of Spanish (third semester) were assigned to one of three instructional treatments: input practice, output practice, and no practice. Subjects performed three tasks: a comprehension test, a production test, and a written narrative of a 1-minute silent video. The results of a repeated measures ANOVA show that both experimental groups (input processing and output processing) significantly improved their scores compared to the control group on the comprehension test. The results of the production and narration task were not affected by experimental instruction. These findings contradict previous studies supporting the pedagogical value of input processing.