This is the second in a series of three papers attempting to quantify the vocabulary of medical research articles. It continues my investigation into a word frequency count derived from a 21,000 word computer corpus of articles drawn from a range of specialized medical fields. The distribution of the most frequent words across the "IMRAD" structure of medical articles is investigated. Several words are found to be distributed differently across the divisions, and it is shown that the way in which a word is used is often dependent on the section in which it is found. These findings have implications not only for teachers and learners of Medical English, but also for those in the wider field of ESP (English for Specific Purposes).