In light of the transition from a labor- to a knowledge-intensive economic system, member countries of the Organization for Economic Cooperation and Development (OECD) seem to have little choice but to increase the efficient allocation of both the quality and flow of knowledge and literacy skills. To reduce risks and counter the factors that otherwise might lead to underinvestment in human capital, a new human capital accounting system is needed that would aim at measuring and recording levels of competency. In particular, it would focus on the acquisition of knowledge and competency relevant to the workplace. A general survey of workers' knowledge and skills offers the possibility of directly testing how these are related to productivity. Such information would begin to create a more solid basis for policy on education and training. To test the relationship between competence and productivity, the samples should be designed to measure both competence and productivity for the same respondents. This implies collecting data in work settings. The primary sampling units should be firms or work establishments. Another desirable feature of the sampling design would be to sample clusters of entire work groups within enterprises. The following challenges must be overcome to accomplish the study design: establish a definition of skill and develop a skill taxonomy; develop reasonably accurate test measures; collect data; and include the time dimension and reference period for measurement. (Contains 43 references.) (YLB)