The production of experimental cirrhosis in the rat, most commonly by multiple doses of carbon tetrachloride (CCl4), is a difficult process with a low yield of 'cirrhosis' of widely varied histology. This is due to an unpredictable variation in the response of the rat liver to CCl4. Using a method of monitoring the body weight change of the rat in response to intragastric CCl4 has produced a high yield (76%) of cirrhosis with 8-10 doses of CCl4. This improved control over liver damage has now made it possible to produce a 'standardized' type of decompensated micronodular cirrhosis. A simple non-invasive method of determining when this point has been reached, using a visual grading of ascites during light halothane/oxygen anaesthesia, is described.