A multi-cohort cluster randomized trial was conducted to estimate effects of rich vocabulary classroom instruction on vocabulary and reading comprehension. A total of 1,232 fourth- and fifth-grade students from 61 classrooms in 24 schools completed the study. Students received instruction in 140 Tier Two vocabulary words featured in two grade-level novels. Teachers were randomly assigned to either rich vocabulary (treatment) or to business as usual (control). Teachers in the treatment condition allotted 30 minutes per day to the intervention for 14 weeks. Hierarchical linear modeling revealed positive, significant treatment effects on distal and proximal measures of vocabulary and comprehension. However, average distal treatment effects were small (approximate d =0.15) compared with proximal effects (approximate d = 1.24). Observations of teachers' language arts instruction indicated that treatment teachers spent significantly more time on vocabulary and less time on comprehension instruction than did teachers in the control condition. Results support the intensity and depth of the instruction for learning the taught corpus of words, and modest transfer to global vocabulary and comprehension.