The theory that increasing cancer incidence rates in developed countries are primarily the consequence of an expanding ageing population and improved diagnostic testing is widely held. In the United Kingdom the proportion of people aged 50 and over has increased by 45% since 1951 and this proportion is set to increase by a further 36% by the year 2031, so the United Kingdom does indeed have an expanding ageing population. However, the increase in cancer incidence affects people across the whole age spectrum. To test the hypothesis that the age of onset of cancer (overall and specific) in England and Wales is decreasing over time we have developed The Cancer Incidence Temporality Index (CITI), which gives a crude measurement of the portion of the population, in which cancer incidence is rising fastest over time: I=(SigmaO(a)/ SigmaE(a))/(SigmaO(a)/SigmaE(a)), where I is the CITI value, O is the observed number of cases and E is the expected number of cases; 'a' and 'b' refer to separate summation ranges for younger and older age groups. Population data and cancer incidence data in England and Wales, 1971-1999 were obtained from the UK Office for National Statistics. The trends in CITI values have been shown graphically for cancer overall and for specific tumour sites. The impact of diagnostic testing is also addressed. The results of this study suggest that the average age of onset of prostate, breast and cervical cancer is temporally decreasing. The study also suggests that for cancer overall the trend for the age of onset of cancer in males has stabilised since 1990 and has started to reverse in females from 1995 despite the expanding ageing population. A similar trend is observed for leukaemias. The CITI analysis for colon cancer shows that the age of onset in both males and females is increasing over time. The trend for ovarian cancer is similar to that for colon cancer. The CITI analysis for NHL in males is similar to that for colon cancer, however, in females the trend stabilised after 1990. The CITI may aid prediction of changes in the age of onset of cancer and thus aid targeted aetiological research. In addition, we suggest the need for a mathematical model, which may measure the changes in the age of onset of cancer in units of time.