Summary.A group of profoundly deaf 15‐year‐old subjects with no other handicap and of average non‐verbal intelligence were given a lip‐reading test. The same test was given to comparable hearing subjects ‘deafened’ by white noise masking. The difference between the groups was not significant. Performance of the deaf subjects (and the hearing) was, however, significantly better when they read the test items from print. The implication is that the relatively poor lip‐reading ability of the deaf subjects was not because of linguist