This study extends research on written corrective feedback (CF) by investigating how timing of CF affects grammar acquisition. Specifically, it examined the relative effects of synchronous and asynchronous CF on the accurate use of the hypothetical conditional structure. Participants were 68 intermediate-level students of English at a university in Japan. Learners from a synchronous CF group (SCF), an asynchronous CF group (ACF), and a comparison group completed 2 writing tasks using Google Docs. The 2 experimental groups received focused direct CF with the following differences: The SCF group received synchronous feedback on grammatical errors during writing tasks, while the ACF learners received feedback after the tasks. Participants revised their texts upon receiving the feedback. The comparison group completed the writing tasks without feedback. Accurate use of the target feature was measured by a set of 3 text reconstruction tasks conducted as pre-, immediate post-, and delayed posttests. The results showed that both experimental groups significantly improved from the pretest to the 2 posttests while the comparison group did not. Overall, however, effect sizes for the posttests indicated that SCF was more effective in improving learners' accuracy with only the SCF group outperforming the comparison group on the delayed posttest.