That the human frontal lobes are particularly vulnerable to age-related deterioration has been frequently invoked as an explanation of functional decline in aging. This "frontal aging hypothesis" is evaluated in this review by examining evidence of selectively reduced frontal lobe function in aging. The frontal aging hypothesis predicts that functions largely dependent on frontal regions would decline in aging, while functions largely independent of frontal lobes would remain relatively spared. The hypothesis further predicts that age-related brain change would selectively impact frontal regions. The literatures on working memory, visuospatial attention, face recognition, and implicit memory were reviewed as exemplars of functions dependent on prefrontal, parietal, temporal and occipitotemporal cortices, respectively, with a view to establishing mediating structures and effects of aging. Age sensitivity was seen both in functions dependent on frontal integrity as well as in functions apparently independent of frontal integrity. Further, although prefrontal areas exhibit age-related decreases in regional volume, blood flow and metabolism, nonfrontal cortical regions undergo similar declines. It is concluded that while the frontal lobes are subject to age-related changes reflected in both behavior and pathology, there is only weak and conflicting evidence that frontal regions are selectively and differentially affected by aging. It is argued that a network-based theory of cognitive aging has advantages over the localizationist approach inherent in the frontal aging hypothesis.