Between 1971 and 1983, Korea's mean gender earnings ration remained virtually stagnant at 47 percent. But after 1983, the earnings ration took a distinct turn upward. In other words, not until after 1983 did Korean women make any progress in closing the gender-earnings gap. when controlling for education, the analysis reveals a surprising drop in relative earning across education groups in the 1970s and early 1980s, and a recover thereafter. The author uses an extremely rich set of microdata (suitable for decomposition) to explain the trends in Korea's earnings differential. Results indicate that most of the 1983 reversal is attributable to a strong compression in market returns to skills and to narrowing gender differences in education and experience. The widening gender earnings differential across education groups before 1983 resulted primarily from a growing gender gap in unobserved characteristics. Growing gender differences in unmeasured ability or increased wage discrimination could explain this trend. After 1983, women with high school education or less benefit primarily from a dramatic narrowing in the economy's distribution of market payoffs to skills, enough for women to begin to catch up to men relative earnings. A compression in the return to skills helped only some groups. Women with college educations did not experience increased benefits from changes in the market payoff to skills. Stricter enforcement of Korea's equal-pay-for-equal-work provision could help reduce the outright discrimination against women workers that might be the underlying problem. By boosting the potential of Korea's female labor force, stronger enforcement of Korea's equal opportunity provisions would improve the country's economic productivity.