The use of memory aids to improve respondent recall is becoming increasingly popular in health surveys. In a one-year field experiment involving over 1,200 respondents, people assigned memory aids reported over 50 per cent more symptom episodes than those not assigned memory aids. Differences in reporting levels were particularly striking during the first six to 12 weeks of the survey, for symptoms related to acute, transitory illness, and among minorities and the lower socioeconomic groups. Respondents assigned memory aids were also more likely to report a doctor contact--especially during the fall/winter months. Among respondent who used their memory aids at least once (N = 252), nearly 35 per cent said they were influenced by the information recorded in their memory aids, and most of those who were influenced said that they became more aware of their health. Ten of these people said that they became more worried or depressed because of the information recorded in their memory aids. This constitutes about a four per cent risk among those people who reported using their memory aids at least once. There was no evidence to suggest that the memory aids influenced perceptions of health status.