IN AN earlier paper, we1 reported the findings in 45 mucosal biopsies from the tympanum, obtained in connection with a radical mastoid operation. Fourteen of the 25 ears without cholesteatoma showed squamous epithelium in the tympanic mucosa. In 12 of 16 ears with clinical cholesteatoma, but no gross extension into the tympanum, there were areas of squamous epithelium in the promontory mucosa. Except for the amount of keratin no difference could be observed between the type of epithelium forming excessive cholesteatoma and the one not associated with cholesteatoma formation. In several areas sharp borderlines were seen between squamous and ciliated epithelium. It was concluded that it is the type of squamous epithelium forming cholesteatoma that needs thorough removal by radical ear surgery whereas such meticulous care may not be needed with that type of squamous epithelium not producing cholesteatoma. In an infected ear, it was assumed, the ciliated epithelium recedes