This paper examines methods for estimating the efficiency of measles and pertussis notification with particular reference to England and Wales. Crude estimates are obtained from a comparison of annual numbers of births and notifications, and this approach is modified to include detailed age-specific data. Other sources of data, notably the Hospital In-Patient Enquiry, are used to give detailed estimates over time. These analyses provide evidence for a strong positive correlation between notification efficiency and incidence for both diseases and for a dramatic fall in the notification efficiency for pertussis between 1957 and 1976. This decrease is of particular interest insofar as it could explain in part the apparent fall in size of successive pertussis epidemics during that period. It is estimated that since 1976 the spotter practices of the Royal College of General Practitioners General Practice Research Unit have been approximately 1.5 times more efficient at reporting measles and 2.5 times more efficient at reporting pertussis than the national average for reporting to the Office of Population Censuses and Surveys. Finally it is noted that estimates of notification efficiency are positively correlated with the assumed level of vaccine efficacy, and that all these methods require assumptions concerning the proportion of vaccinations which are recorded, the proportion of vaccinations which are successful in immunizing the recipient, and the proportion of the population which ultimately contracts the infection. In general, measles notification in England and Wales is considerably better than that for pertussis, being of the order of 40-60% for the former and only 5-25% for the latter.