The work of many researchers has firmly established that random, non-enzymic crosslinking occurs in aging. Aluminium is one of the most powerful, and of the most widely distributed crosslinking agents. The published information on aluminium content in human nutrition has given highly divergent results. Since aluminium has been implicated in human neural and other disease, we have undertaken a broad study of aluminium content of food, using throughout the study atomic absorption spectrometry. All of the data reported have been obtained with the same methods and instrumentation set aside expressly for only this project. All determinations were made in triplicate, by the same researchers. If results differed more than 15% a rerun was made. The results show differences of up to over a hundred times, and commonly over 10 times between the highest and the lowest values of any foods analysed. These data show lack of control. Findings also indicate that if only the top 10% could be avoided, from a health standpoint significant reduction of aluminium intake would be achieved.