This article reports on a classroom-based study that examined the effects of direct focused and unfocused written corrective feedback (WCF) on the accuracy development of the past hypothetical conditional (e.g., "If I had had money, I would have bought a new computer.") and the indefinite article (a/an). The study employed a pretest-posttest-delayed posttest design with two feedback treatment sessions. Six intermediate classes, a total of 108 college freshmen at a university in Korea and two native English speaking instructors, participated in the study. The classes were assigned into a focused WCF group, an unfocused WCF group, and a control group. A text-reconstruction task was used as a writing material and a test instrument. The results revealed that both focused and unfocused WCF groups outperformed the control group in the accuracy gains of both target forms. When the relative effects of two types of feedback were considered, there was no difference in the contribution focused and unfocused WCF made for the accuracy improvement of the hypothetical conditional. However, focused WCF was more beneficial than unfocused WCF for the development of the indefinite article.