A study examined how the occupational structure of the Australian labor market evolved and how individuals fared in the process. It identified issues in defining skill and knowledge and followed Elias and McKnight (2001) in stating that sufficient evidence showed a very high correlation between job-required cognitive ability and ordinal skill ranking. Prong 1 of an empirical approach examined census data on occupational composition of employment from 1986-2000 and showed that employment grew most rapidly in professional jobs and intermediate clerical, service, and sales jobs, and a very large number of trades were in decline. Prong 2 examined longitudinal data from the 1997 Negotiating the Life Course Survey with work and education histories for over 2,000 people and found that about six in seven changed occupation between their first main job on entering the labor market and their present job, and just over half changed broad skill ranking. The most important determinant of whether a person began working life in a knowledge job and stayed was education. There was little association between people's background characteristics, education, and work experience and whether they moved into knowledge work. Implications for vocational education and training (VET) were that the surest path to knowledge jobs is to obtain post-school qualifications; VET might deliver degree-level courses at the associate professional level where diplomas are often required; and over-education through VET is dangerous if educational attainment outstrips growth of jobs at the top of skill distribution. (Contains 28 references) (YLB)