This paper examines the problems involved in subject retrieval from full-text databases of secondary materials in the humanities. Ten such databases were studied and their search functionality evaluated, focusing on factors such as Boolean operators, document surrogates, limiting by subject area, proximity operators, phrase searching, wildcards, weighting of search terms, limiting by type of document, controlled vocabulary indexing and ranking, and display of search results. The author suggests ways in which full-text searching might be improved, whether by enhancement of database records, by introduction of enhanced search functionality, or by the education of searchers in more effective search techniques. The conclusion is that current digitization projects are not producing databases that meet the needs of scholars. (Contains 28 notes.)