For the assessment of fat malabsorption, the standard method of measuring faecal fat excretion using a 5 day stool collection has been compared with the alternative methods: stool microscopy, a lipid tolerance test and a continuous marker technique for the estimation of fat content on a single stool sample. The lipid test, using an emulsion of arachis oil (Prosparol), was less reliable than had been expected with a sensitivity of 33% and a specificity of 45.4%. Stool microscopy using Oil Red O to stain fat globules had a sensitivity of 72.2% and a specificity of 95.4%. Fat estimation of a single stool sample using copper (1) thiocyanate showed a high correlation with that determined on a 5 day stool collection (p less than 0.001). It is concluded that lipid tolerance tests have little place in the estimation of fat absorption. In laboratories where faecal fats are not measured, microscopic examination of stool for fat globules provides a specific and relatively sensitive method for detecting steatorrhoea. The use of a continuous marker provides a method for assessing the degree of steatorrhoea on a single stool sample without the disadvantages of the conventional method of faecal fat analysis.