This article reports on a classroom-based experiment that tested the effects of three vocabulary teaching constructs (involvement load, spacing, and intentionality) on the learning of English verb-noun collocations--for example, "shelve a plan." Laufer and Hulstijn's (2001) "involvement load" predicts that the higher the motivational-cognitive load of a task, the more effectively it promotes word retention. "Spacing" refers to the advantage of spreading out learning opportunities for words as opposed to massing them. "Intentionality" comprises two word processing modes: intentional learning (posttest announced) and incidental learning (posttest unannounced), where the former is claimed to outperform the latter. The constructs were integrated into an intervention study with 59 adolescent L1 Swedish learners of English in within- and between-subjects designs. Learners processed target items three times when performing tasks that operationalized the constructs. Three posttests of productive knowledge of target items were administered. Statistical analyses of gain scores show that neither involvement load nor spacing had a significant positive impact on learning gains. Significant effects were found on three measures for intentional learning when compared to incidental learning. The findings are discussed in relation to previous research and their implications for English language teaching (ELT).