A computerised diagnosis, AGECAT, is applied to data from random community samples of the elderly in New York and London in order to examine the prevalence of mental illness in the two cities, especially the result reported from the same study, using a different diagnostic procedure, that the proportion of dementia was higher in New York. The greater proportion of dementia in New York is confirmed for all AGECAT's levels of diagnostic confidence, for both sexes and at all half-decades over 65 years. Provided that this difference is upheld and genetic differences can be discounted, these findings would seem to support a contributory environmental cause for dementia. Proportions of depression as a whole are similar in both cities. Syndrome case and sub-case levels of neurotic disorder are generally higher in London.