A mathematical model of age exaggeration is suggested that enables studying the statistical manifestations of the age overstatement formally, in simulations, and empirically. Our findings indicate that even an infrequent age exaggeration may substantially distort the mortality curve at an advanced old age. Old-age mortality distortions that were previously attributed to the effects of mortality selection in the heterogeneous population are shown to be produced by the age exaggeration too. The model proposed here may be used in both reconstructing the actual mortality profile and assessing the mode and extent of age misreporting.