The dog chew studied here is a starch-based, twin-screw cooker extruded dog care and treat (C&T) product with oral health benefits. The manufacturing process and nutrient profile of such products are markedly different from those of main meal pet foods. Predicted metabolisable energy (PME) in pet food is calculated using equations derived from main meal feeding studies so it is unclear whether these equations can be applied to C&T products. The present study aimed to directly measure metabolisable energy (ME) content of the dog chew in dogs and compare with calculated PME. A batch of dog chews was manufactured and the product rendered micronutrient complete to allow solus feeding. Following a 3 d standard diet pre-feed phase, the test product was fed solus to a panel of seven adult dogs for a period of 8 d. Dietary intake was recorded daily and faecal matter collected for the last 5 d. Test product and pooled faecal samples were analysed for proximate nutrients, and digestibility coefficients were calculated as the difference in intake and faecal excretion (7–11 d). Digestible energy was converted to ME by correcting for energy losses in urine. PME was calculated using proximate analysis and modified Atwater factors according to National Research Council 2006. The results showed close agreement between actual ME (1272 (sd 12·1) kJ/100 g) and calculated PME (1268 (sd 12·6) kJ/100 g), indicating transferability of the NRC 2006 PME equations to the dog chew tested here.