This paper discusses the conceptual, definitional and operational problems which result in under-enumeration of female workers in large-scale censuses and sample surveys. A detailed examination of the definitions used and the estimates obtained reveal that certain Indian censuses and NSS rounds had a better enumeration of women workers than others. Taking the comparable censuses/NSS rounds over the period 1961 to 1983, the widely alleged decline in female participation rates is seen to be not substantiated. The recent decades also show a shift in the workforce away from agriculture and a trend towards increasing casualisation of the workforce in the rural areas.