At least one third of all U.S. college students don't get a degree, even six years after they enroll. Earlier research focusing on trends through 1990 found broad declines in college graduation rates, especially among men attending less-selective four-year schools. Since then, however, the picture of college enrollment has changed dramatically, with increases in both high-school graduation and college matriculation rates. Have completion rates changed as well? To find out, the authors analyzed federal education and Census data and found that rates of college completion have gone up since 1990, at a broad mix of institution types and among both men and women. This trend is confirmed in federal data, the Census, and registrar data from 10 public universities. What's driving this growth? The authors looked at student background and academic preparation, as well as institutional practices like support-service spending, and find that none of these potential factors explain the changes. But one trend is clear across all the datasets: compared to decades past, college students have been earning better grades in recent years, and better college grades are strongly associated with higher rates of graduation. The authors explore a range of factors that could influence student performance, such as high-school preparation and rates of labor-force participation in college and find that these would predict students to be less likely to graduate, not more. While earning a degree is something to be lauded, evidence of grade inflation raises important questions about the meaning of some college degrees. And with growing uncertainty about the relative return on investment in terms of both time and tuition, both students and institutions should take a hard look at the ultimate value of their efforts.