Studies examining decontextualized associative vocabulary learning have shown that long spacing between encounters with an item facilitates learning more than short or no spacing, a phenomenon known as distributed practice effect. However, the effect of spacing on learning words in context is less researched and the results, so far, are inconsistent. In this study, we compared the effect of massed and spaced distributions on second language vocabulary learning from reading. Japanese speakers of English encountered 48 novel vocabulary items embedded in informative English sentences, inferred their meanings from contexts, and received feedback in the form of English synonyms and Japanese translation equivalents. To test the hypothesis that the effects of spacing might differentially affect the development of explicit or tacit word knowledge, spacing effects were measured using semantic priming as well as a meaning recall and a meaning-form matching posttest. Results showed an advantage of spaced over massed learning on the meaning recall and meaning-form matching posttests. However, a similar semantic priming effect was observed irrespective of whether an item was encountered in the massed or spaced distribution. These results suggest that the spacing effect holds in contextual word learning for the development of explicit vocabulary knowledge, but massing appears to be as effective as spacing for the acquisition of tacit semantic knowledge.