The purpose of this study was to investigate the efficacy of an 18-week program of direct and extended vocabulary instruction with kindergarten students on both proximal measures of target word knowledge and transfer measures of generalized language and literacy. A second purpose was to examine whether treatment effects would be moderated by initial receptive vocabulary knowledge measured at pretest. In a quasi-experimental design, 80 kindergarten students from schools serving large at-risk populations were taught the meanings of 54 vocabulary words within interactive story read alouds over 36 half-hour instructional lessons (2 lessons per week over 18 weeks). An additional 44 students served as a no-treatment control. Findings indicated that students who received vocabulary instruction outperformed controls on a measure of target word knowledge as well as measures of generalized receptive vocabulary and listening comprehension. In addition, initial receptive vocabulary was strongly related to posttest performance on all measures. Implications are discussed in relation to supporting vocabulary development in the early grades within a multitier framework of instruction and intervention. (Contains 10 tables and 4 figures.)