Abstract On the basis of data collected in the 1975 National Fertility Study (NFS) in which continuously married white women interviewed in the 1970 NFS were reinterviewed five years later, the aggregate and individual validity of reported fertility intentions is evaluated. The main finding about aggregate prediction is that 1970 intentions overestimated 1971–1975 fertility to the same extent as the conventional 1970 total fertility rate. The conclusion is that intentions suffer all of the same vulnerabilities as other period measures. At the individual level, the validity of intentions is considerable in comparison with most other indicators.