Background: Vocabulary scores increase until approximately age 65 years and then remain stable or decrease slightly, unlike scores on tests of other cognitive abilities that decline significantly with age. Aims: To review the findings on ageing-related changes in vocabulary, and to discuss four methodological issues: research design; test type; measurement; and vocabulary ability as a proxy for general intelligence. Main Contribution: A discussion of cross-sectional and longitudinal research designs shows that cohort membership accounts for some but not all ageing-related changes in vocabulary, and that drop-out and test-retest effects do not alter conclusions regarding these changes. Test type affects age trends in vocabulary, and if researchers use only one test, they should choose a multiple-choice synonym test. While some authors suggest that vocabulary tests do not measure the same underlying ability in younger and older adults, more research of this suggestion is needed. A brief examination of the use of vocabulary ability as a proxy for general intelligence in healthy ageing and for premorbid abilities in dementia indicates that such practice is often questionable. Conclusions: Vocabulary knowledge increases through the mid-60s regardless of measurement method. However, there is little information on how word knowledge serves other verbal skills in old age, how and when adults learn new words, or how much exposure is necessary for meanings to remain in storage for a lifetime. Research of these issues may require new methodologies, as well as novel theoretical accounts of ageing-related effects on vocabulary.