Wood, K. M., Montanholi, Y. R., Fitzsimmons, C. F., Miller, S. P., McBride, B. W. and Swanson, K. C. 2014. Characterization and evaluation of residual feed intake measured in mid- to late-gestation mature beef cows and relationships with circulating serum metabolites and linear body measurements. Can. J. Anim. Sci. 94: 499-508. To evaluate the use of residual feed intake (RFI) models measured in mid-to-late gestating beef cows, a dataset was used combining data from five experiments containing nine treatment and/or replicate groups for a total of 321 animal records. Investigations of RFI models included the effects of age, ultrasound measures of body composition, pregnancy corrected gain, and dietary treatment group. A subset of animals had serum analyzed for circulating metabolites (glucose, non-esterified fatty acids, urea, beta-hydroxybutyrate; n=227) and/or linear body measures (hip height and width, body length, body circumference at the heart, mid-body and at flank; n=114) for correlation analysis with efficiency traits. Goodness-of-fit for all RFI models was assessed using R², CV, and Bayesian information criteria. Across treatment/replicate groups, the largest improvements in model fit were made by accounting for management group and dietary treatment. Circulating urea concentrations were positively correlated (P ≤0.05) with average daily gain, dry matter intake, gain:feed, and group RFI model. Linear body measures and circulating metabolites measured (with the exception of urea) were not correlated (P>0.05) with economically relevant traits. Measures of RFI as an indication of feed efficiency may be challenging for use in gestating beef cows. Large variation in efficiency between cows remain, and may be related to mechanisms influencing maintenance and energy expenditures.