This study examined the effects of Reading Intervention for Adolescents, a 2-year extensive reading intervention targeting current and former English learners identified as struggling readers based on their performance on the state accountability assessment. Students who enrolled at three participating urban high schools were randomly assigned to the Reading Intervention for Adolescents treatment condition (n = 175) or a business-as-usual comparison condition. Students assigned to the treatment condition participated in the intervention for approximately 50 minutes daily for 2 school years in lieu of a school-provided elective course, which business-as-usual students took consistent with typical scheduling. Findings revealed significant effects for the treatment condition on sentence-level fluency and comprehension (g = 0.18) and on a proximal measure of vocabulary learning (g = 0.41), but not on standardized measures of word reading, vocabulary, or reading comprehension (g range: -0.09 to 0.06). Posthoc moderation analyses investigated whether initial proficiency levels interacted with treatment effects. On sentence-level fluency and comprehension and on vocabulary learning, initial scores were significantly associated with treatment effects--however, in opposite directions. Students who scored low at baseline on sentence reading and comprehension scored relatively higher at posttest on that measure, whereas students who scored high at baseline on the proximal vocabulary measure scored relatively higher at posttest on that measure. The discussion focuses on the difficulty of remediating persistent reading difficulties in high school, particularly among English learners, who are often still in the process of acquiring academic proficiency in English. [This paper was published in the "Journal of Educational Psychology," 2018.]