Use of Shikata's stain was found to be a sensitive method for the demonstration of copper-protein complexes, but use of the rhodamine stain may be more reliable. Interobserver agreement was slightly better with Shikata's stain. Hepatic copper increased with the histologic progression of chronic nonsuppurative destructive cholangitis (CNDC). All specimens that contained Mallory bodies and most specimens (91%) that contained bile yielded positive copper stains. No correlation was found between positive copper stains and other histologic features. A strongly positive copper stain aided in the diagnosis of CNDC. All specimens with negative copper stains contained less than 250 microgram of copper per gram of dry weight. From a strongly positive copper stain, the chemical copper content could not be predicted with certainty. Shikata's method seemed adequate to evaluate penicillamine treatment trials.