The objectives of this dissertation were to: 1) develop and validate equations used to estimate individual cow dry matter intake (DMI; kg/d) based on a nitrogen (N) balance approach, 2) determine the discriminatory power of several biological, production, and dietary variables on dairy feed efficiency (FE) as defined as energy-corrected milk (ECM; kg/d) per unit of DMI, 3) repeat the second objective using residual feed intake (RFI) to indicate FE status, and 4) determine if RFI values are dependent on the equation utilized to estimate DMI.

Results from the first experiment (Chapter 3) indicated that DMI could be successfully estimated on an individual cow basis using the following commonly measured parameters: milk yield, milk protein concentration, body weight (BW; kg), and dietary N concentration. These inputs are relatively simple to measure; therefore, this equation may be used in the dairy industry as a practical method to estimate individual cow DMI when cows are fed in a group setting.

The results of the second experiment (Chapter 4) suggested that days in milk (DIM), milk fat yield (g/d), and BW had the most discriminatory power (89% success rate) to discriminate between cows based on their FE status when FE was defined as ECM per unit of DMI. Therefore, dairy producers can use these 3 variables to select for cows with high FE without requiring the measurement of DMI which can be costly and difficult to obtain.

Observations from the third experiment (Chapter 5) suggested that RFI is indicative of differences in metabolic efficiency between cows independent of most biological, production, and dietary variables, except DIM. These results are consistent with other studies that have suggested that RFI is indicative of true differences in metabolic efficiency between cows regardless of production parameters.

Lastly, the results of the fourth experiment (Chapter 6) suggest that RFI values generated from different DMI equations are strongly correlated such that RFI values are independent of the DMI equation utilized in the calculation. Thus, dairy producers can select the equation to estimate DMI that is most suitable for their operation without causing an “equation bias” on the RFI calculation.