Previously reported concentrations of triacylglycerol in skeletal muscle have shown high coefficients of variation, and there have been large differences between mean concentrations reported in a given muscle. Conditions for sampling and measurement were therefore investigated. Samples were best taken under anesthesia as breakdown of triacylglycerol was rapid after decapitation. Silicic acid was preferable to zeolite for removal of phospholipids although either agent could interfere with the estimation. Even with apparently reliable methods, a high variability was found in any one muscle and there were large differences between muscles. It is unlikely that the variability was due to contamination with adipose tissue. Concentrations of glycogen and phospholipid were much less variable. Although the store of triacylglycerol in skeletal muscle in caloric terms was found to be 2-18 times greater than that of glycogen, the variability found is likely to hamper studies of its metabolic role.