“Faux amis” or “false friends” (abbreviated hereafter as FF) are metalinguistic terms used to name word pairs that look alike in two languages, but that do not have the same meaning. The purpose of this paper is to identify and classify FF in a corpus of medical English prose. Our source material consists of twenty research papers randomly chosen and published between 1994 and 1996 in various leading medical journals. Every FF was identified by means of a contextual analysis. The deceptive cognates recorded were classified according to two groups of variables: FF types and word classes. The proportion of FF found in the corpus reached 5.3% of the number of total running words. This leads us to believe that they deserve special attention since it is well known that they represent an important problem for terminologists and are frequently misinterpreted by medical doctors and students. We thus believe that the elaboration of tables containing word pairs (English-Spanish, in our case) most frequently encountered, along with well designed exercises, could prove most helpful if presented at an early stage of reading comprehension and translation courses.