The objective of this study was to investigate the association of residual feed intake (RFI) with routinely measured milk components that are indicators of subclinical mastitis and ketosis. Milk somatic cell count (SCC, 10³ cells/mL) is commonly used to diagnose subclinical mastitis while β-hydroxybutyrate (BHB, mmol/L) and acetone (ACT, mmol/L) are indicators of ketosis. RFI was phenotyped in 71 lactating Holstein dairy cows at the Dairy Research and Technology Center–University of Alberta with components of metabolic body weight, empty body weight change, and milk production energy requirements over 255 d in milk using random regression and multiple linear regression models. Correspondingly, test-day milk samples were collected twice a week and analyzed at DHI lab by a MIR spectrometer (MilkoScan FT+, Foss, Hillerød, Denmark) during the same period as for RFI prediction. A total of 3,810 test day records for each of the traits; SCC, BHB and ACT were obtained from April to August 2016. The data were analyzed using a MIXED model procedure of SAS with fixed effects of RFI (-RFI and +RFI), lactation number (1, 2 and 3+), milking time (AM and PM), interactions of RFI x lactation, RFI x milking time and random effects of cow. Days in milk was included in the model as a covariate. The result indicated that -RFI and +RFI groups did not differ in SCC (381.01 ± 55.77 vs. 359.47 ± 47.14; P = 0.76), BHB (0.53 ± 0.07 vs. 0.64 ± 0.05; P = 0.25) and ACT (0.30 ± 0.06 vs. 0.32 ± 0.04; P = 0.75). The correlation analysis also showed no evidence of RFI association with SCC (r = 0.01; P = 0.91), BHB (r = 0.17; P = 0.17) and ACT (r = −0.042; P = 0.72). The result suggests that selection for RFI may not be negatively correlated with incidence of subclinical mastitis or ketosis in dairy cattle. Estimation of the genetic correlations of RFI with SCC, BHB, and ACT in a larger sample is warranted to confirm these preliminary results.