Abstract The objective of the present study was to develop prediction equations for digestible energy (DE) and metabolizable energy (ME) in swine diets based on chemical compositions of feces with no dietary information. A total of 262 fecal samples obtained from growing pigs fed 27 diets from 4 experiments were used. The diets were based mainly on corn, soybean meal, and other plant-originated feed ingredients. In the digestibility experiments, feces and urine were quantitatively collected using the marker-to-marker procedure. The diet and fecal samples were analyzed for dry matter (DM), crude protein, ash, and neutral detergent fiber (NDF). Digestible energy in the 27 diets ranged from 2,907 to 4,374 kcal/kg DM and ME ranged from 2,774 to 3,983 kcal/kg DM. Fecal ash was positively correlated with DE (r = 0.81, P < 0.001) and ME (r = 0.66, P < 0.001) whereas fecal NDF was negatively correlated with DE (r = –0.66, P < 0.001) and ME (r = –0.74, P < 0.001). The energy prediction equations developed using fecal chemical components were: energy digestibility = 88.1 + 0.37 × ash – 0.28 × NDF (R2 = 0.57, P < 0.001); DE (kcal/kg DM) = 3,604 + 32.8 × ash – 11.1 × NDF (R2 = 0.71, P < 0.001); and ME (kcal/kg DM) = 4,077 + 15.7 × ash – 18.1 × NDF (R2 = 0.63, P < 0.001) where chemical components are expressed as % DM. In conclusion, energy digestibility, DE values, and ME values in swine diets are correlated with fecal ash and NDF concentrations and the energy values can be predicted using fecal ash and NDF concentrations without dietary information.