In the initial stage of lactation, dairy cattle experience negative energy balance, as the energy expenditure of milk production exceeds the amount of energy the cow is able to take up through eating. In this state, there is increased body fat mobilization by the cow, with consequences in milk composition, particularly in fatty acid profile and possibly increased ketone bodies in both blood and milk. On-farm technologies allow novel monitoring techniques and increase the overall amount of data available to producers and researchers. The objective of this study is to analyze the pallet of automatically collected data sources currently recorded at the University of Guelph’s Livestock Research and Innovation Centre (LRIC) and relate them to changes in milk composition. The automatically collected data sources currently recorded at LRIC include, but are not limited to, automatically recorded cow weights via a walkover parlour scale, automatic body condition scoring via 3-dimensional imaging, as well as individual feed intake events. Moreover, milk testing is performed weekly, with the mid-infrared spectral data also obtained for every test. Major milk fatty acid groups were predicted using previously developed calibration equations, and predicted milk beta-hydroxybutyrate contents also obtained. Changes in cow body weight in the initial stage of lactation were examined, along with milk composition. The results of this study improve the understanding of the phenotypic relationship between body weight changes and milk composition in early-lactation dairy cattle. Furthermore, we explore phenotypes associated with energy balance with implications for cow health and future genetic selection strategies.