It is known that speed and accuracy in recognizing words are constrained by their frequency of occurrence ("frequency effect"). This study bears on the diachrony of the word frequency effect. Our postulates in this respect were (1) that a significant frequency effect should be present throughout adulthood, irrespective of age, and (2) that the magnitude of this effect should be greater among the elderly. Twenty young and 20 older healthy adults were submitted to a lexical decision experiment. Results confirmed our first postulate but invalidated the second one, that is, significant frequency effects were found in both groups but these effects were documented to be of identical magnitude. An attempt is made at explaining the latter result from a theoretical standpoint. The former is interpreted as further evidence that senescence (normal aging) does not interfere with passive, automatic and unconscious mental processes. Moreover, it is suggested that-if observed among otherwise apparently healthy elderly adults-modifications of the frequency effect might be taken as a cognitive marker of disease.