This thesis research was aimed to develop a residual feed intake (RFI) prediction equation in dairy cattle while accounting for the animals’ multifunctional lactation non-linear energy requirements profiles. The possibility of shortening RFI test period and finding indicator traits for RFI selection were also investigated. A total of 281 first-lactation dairy cows at the Dairy Research and Technology Center of the University of Alberta from June 2007 through October 2012 were used. Individual daily feed intakes, repeated measurements of monthly body weight, and body condition scores of these animals were recorded from 5 to 305 days-in-milk. Milk production and milk composition data were extracted from the Dairy Herd Improvement Program, and their first type classification data was retrieved from the Canadian Dairy Network. To reduce the test period in a lactation, the acquired data from whole lactation (5 to 305 DIM) was subdivided equally into three shortened periods, early, mid, and late stage of lactation. RFI prediction equations were developed for the whole and each of the shortened periods. Each animal, based on its predicted RFI value, was assigned to high (RFI> 0.5 SD) medium (RFI= ± 0.5 SD) or low (RFI< 0.5 SD) RFI classes within each of the test period. Compared with whole lactation, numbers of the animals’ remaining in the same RFI class within any of the shortened test periods were determined to study the consistency of RFI prediction. Moreover, genetic and phenotypic correlations between conformation traits and RFI were estimated to investigate the possibility to use these conformation traits as indicator traits for RFI selection. The results showed that RFI could be predicted in whole, early, mid and late lactation with R-square of 0.68, 0.47, 0.49, and 0.79, respectively. Compared with the whole lactation, most of the animals (65.5%) remained in the same RFI classes in mid stage, so mid RFI prediction could be considered as the best representative of whole RFI in compare with early and late periods. Moreover, combinations of eight conformation traits could be used as indicator traits for RFI selection, since they had high genetic correlation with whole lactation RFI.