Summary The term “knowledge worker” has entered the language of management and economics, becoming popular or even fashionable. Consequently, its definitions are varied and often ambiguous or provisional – which makes it difficult to identify and research such employees. Deeper understanding of the term “knowledge workers” is required. One of the ways of defining a term is to define its semantic area through borders and overlaps with synonyms and “adjoining” terms. Such comparisons can help to deepen understanding of the central term and reveal its defining features. Therefore, two objectives were set in this study: 1) to compare “knowledge workers” and synonymous and associated terms specifying similarities, differences and areas of overlap, in order to find defining features of “knowledge workers”; and 2) to specify the synonymous and associated terms closest to knowledge workers and useful as proxies for research purposes. A group of 15 synonyms was selected, including historical and presently applied terms, proposed by various researchers or used in popular language. These terms were first characterized, and then compared to knowledge workers in terms of similarities, differences and areas of overlap. Comparison pointed to a number of features strongly related to knowledge work. Based on them, a sketch definition was proposed: Knowledge workers work mainly on symbols (representations), transforming them in cognitive processes, which is the main source of added value. To do that, they must command a large body of knowledge equivalent to university education, understood and internalised, grounded in experience and consequently updated. They perform complex tasks, focus on problem-solving, creating knowledge, distributing it and applying to achieve results. They broadly use documents and ICT, and require high level of autonomy. With respect to the second aim of the study, there are several terms closer in meaning to knowledge workers. Terms with most similarities and least differences include specialists and experts. Of these, specialists have most similarities, while experts are a narrower, more advanced type of specialists. In terms of overlap, categories closest to knowledge workers seem to be specialists/ professionals (wider) and professionals (narrower) – both are large sub-sets of knowledge workers, with specialists leaving fewer remainders. As far as identifiability is concerned, terms related to classifications of occupations are those better defined. Among them, “white-collars”, knowledge producers and distributors, and information workers were measured in older classifications, while specialists/ professionals and managers are categories measured today. Both are defined in detail in ISCO-08, making them identifiable also at the organisational level. Statistics of specialists and managers are gathered regularly in most countries belonging to ILO and are easily accessible. Concluding, specialists/ professionals (wider) are the best proxy group for researching knowledge workers: they possess almost the same key features, overlap closely, leaving smallest remainder, and are defined in detail in an international classification used worldwide to measure their national populations. Selecting a random member of the group, one is practically certain to find a knowledge worker.