We were interested to read the article by Knox on the new approach to using local authority data in addressing the effects of air pollution on health. To our knowledge, this has been done only once before and does have the benefit of taking into consideration other confounding variables which time-series studies and cohort studies are often unable to do. However, there are a number of problems with this paper that make the data very difficult to interpret, and almost certainly result in substantial overestimates of the true health effects. In addition, the mechanistic interpretation put forward by the author does not cohere with current knowledge of the effects of air pollution on health.