This study examined metabolic profiles of commercial dairy herds differing in milk production.Thirty herds were selected with 10 herds averaging more than 7258 kg milk, 10 herds between 6350 and 7258 kg, and 10 herds averaged below 6350 kg.In each herd, 28 cows were selected for blood sampling, which was collected every 3 mo during 12 mo.Cows varied in estimated relative producing ability.Metabolic profiles were established for 19 blood components.Differences among treatment means were significant between high and low producing herds and little between profiles for milking and pregnant nonlactating cows.Metabolic profiles of problem herds remained within normal ranges for all herds except one.Profiles of herds with breeding or mastitis problems remained in the normal range.Metabolic profile testing was of limited value in screening a dairy herd for potential problems or deficiencies.

Computerized least-cost rations were simulated for 15 dairy herds from herd-average body weights either reported by Dairy Herd Improvement or by taping heart girths of cows. Differences between simulated rations were predicted for feed costs, intakes of crude protein and net energy, and gain in body weight. Weights reported by Dairy Herd Improvement were 68±37kg per cow less than taped weights. The 360 taped cows averaged 633±39kg body weight. The use of reported mean body weights in the formulation of rations for the milking herd increased predicted intakes of net energy and gains in body weight. The effect upon predicted feed costs and intakes of crude protein was not significant. To compensate for differences in estimations of body weight does not appear to be practical. The error in estimation of body weight is small as compared with other possible errors which affect management decisions.