The existence and extent of cyclically-induced changes in the size of the conventionally-defined labour force is examined for women aged 20-59 in Britain, 1951-81. The method is to compare changes in the actual labour force between decennial censuses and recent biennial Labour Force Surveys with estimates of the hypothetical change in the labour force had the pressure of demand remained constant. 'Actual' changes in the labour force have been adjusted for changes in census coverage. The 'hypothetical' changes have been simulated from an econometric model fitted to time series data on the proportion of women in each age group or cohort who are employees or students. These methods are not precise, but if there were large and consistent 'discouraged worker' effects among prime-age women we would have expected stronger evidence of the phenomenon than this data provided. Whether or not potential workers are counted as members of the labour force seems to be far more sensitive to the way in which evidence about the unregistered margins of the labour force is collected than to the availability of job opportunities.