Remote, automated data collection may assist in the goal of precision feeding to reduce variation between individual animals. The objective of this study was to evaluate variation associated with integration of 3 different systems, each estimating different parts of the nutritive energy balance in dairy cows. Twenty-nine Holstein cows, 9 primiparous and 20 multiparous, ranging from 22 to 472 DIM were housed in a freestall pen and subjected to automatic data collection for a period of one month. Milk yield in a parlor was recorded at 0530 and 1730 h daily. Milk samples were collected 4 times weekly and analyzed for fat, protein and lactose content. Body weight and BCS were recorded twice daily using an automated scale and 3D imaging camera (DeLaval, Tumba, Sweden) as cows exited the parlor. Intake of a partial mixed ration based on corn and alfalfa silages was recorded continuously using Insentec feeders (Hokofarm, Marknesse, the Netherlands). Respiratory emissions of CO₂ and consumption of O₂ were recorded with a GreenFeed system (C-Lock Inc., Rapid City, USA), programmed to offer individual cows up to 440 g per visit of a pelleted protein supplement up to 12 times daily. Cows visited the GreenFeed 4.06 ± 1.82 (mean ± SD) times per day, with an average duration of 6.23 ± 1.05 min, for a total of 21.1 ± 9.9 min/d of recorded gas exchange. Diet and milk energy flows were estimated according to NRC (2001). Heat production was estimated from respiratory gas exchange. Variation in observations due to fixed effects of cow and day was estimated by 2-way ANOVA using PROC GLM of SAS (SAS software version 9.4; SAS Institute Inc., Cary, NC). NEL intake was 39.1 ± 5.4 Mcal/d, milk NEL output was 28.1 ± 7.1 Mcal/d, and heat production was 35.7 ± 3.8 Mcal/d. Root mean square errors due to cow and day, respectively, were 57.9 and 25.4% of the mean for NEL intake, 129.2 and 18.6% of the mean for milk NEL output, and 39.2 and 18.7% of the mean for heat production. Variation within cow was less than that between cows. Heat production was the least variable of the 3 energy flows. The integrated systems offer the potential to quantify energy balance in free-moving dairy cows.