Distributed practice is often cited as a best practice for instruction, although research on the application of spaced practice to schools is limited. The impact of practice distributed across days and the impact of practice distributed within sessions were examined. Second and third grade participants (n = 88) were randomly assigned to practice eight spelling words in a massed or distributed fashion, both within sessions and across sessions. Retention was measured immediately and 28 days after the final practice sessions. No differences in immediate retention were observed. Retention after 28 days was greater when across-session distributed practice was provided. These results suggested that practice distributed across weekly sessions was most beneficial.