This dissertation studies the labor market outcomes of college graduates. The choice of college major is one of the most direct ways for college graduates to acquire skills and earnings differences between college majors rival those between high school and college graduates. In each chapter, I combine rigorous descriptive analysis methods with rich administrative data, job postings data and survey data, to study the determinants of earnings differences between college majors. The first chapter studies whether earnings growth and job changing varies with the specificity of a college major's skills to particular occupations. College majors vary in the depth and breadth of human capital, and some majors are associated with skills that are very specific to particular work settings. I find that specific majors initially earn 18% more than general majors, but general majors experience steeper earnings growth, and within thirteen years the earnings premium of specific majors shrinks to 6%. To analyze job changing patterns, I create a new linkage of two nationally representative surveys and administrative earnings records. I find that general majors switch occupations, employers and industries at least 20-30% more often that specific majors. A decomposition exercise suggests that these differences in employer and industry change rates account for 30-50% of the explainable portion of the earnings-growth difference between majors. In the second chapter -- joint work with Brad Hershbein (Upjohn Institute), Steve Hemelt (UNC Chapel-Hill) and Kevin Stange (University of Michigan) -- we document the skill content of college majors as perceived by employers and expressed in the near universe of U.S. online job ads. Social and organizational skills are general in that they are sought by employers of almost all college majors, whereas other skills are more specialized. In turn, general majors -- Business and General Engineering -- have skill profiles similar to all majors; Nursing and Education are specialized. These cross-major differences in skill profiles explain considerable wage variation, with little role for within-major differences in skills across areas. We conclude that college majors can generally be conceptualized as bundles of aggregate skills that are fairly portable across areas in ways that occupations are not. In the third chapter, I investigate the extent to which mean earnings differences between college majors can be accounted for by mean differences in a major's typical job attributes. Using data from the National Survey of College Graduates, I find that accounting for mean differences between majors in job attributes -- including occupation, employer attributes, job tasks and job levels -- accounts for (statistically) over half of the earnings differences between majors. I then use the Gelbach (2016) decomposition to quantify the contribution of each type of job attribute. Unsurprisingly, over one-third of the overall explained between-majors earnings differences is due to occupation. An additional one-third is accounted for by mean differences in employer attributes (ownership structure and size) and an additional 20% is due to variation in job-specific tasks. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]