Data from the University of New Hampshire were used to analyze dry matter intake and milk production in Holstein dairy cows. Equations predicting 4% FCM and dry matter intake were derived from this data using ordinary least squares. Days in milk, available nutrients, calving date, and previous period 4% FCM accounted for 90% of the variation in 4% FCM for heifers and 93% of the variation in 4% FCM for cows with previous lactations. Days in milk, acid detergent fiber, calving date, and dry matter intake from the previous 28 day period accounted for 71% of the variation in heifer dry matter intake and 79% of the variation in dry matter intake for the older cows. First period dry matter intake was predicted from first period 4% FCM, cow body weight, and acid detergent fiber of the ration. The lower R² for the first period dry matter intake (41% for heifers and 43% for cows with previous lactations) indicated that dry matter intake for a cow within a lactation was much more predictable than dry matter intake between different cows. Rations were balanced for absolute amounts for simulated lactating cows using NRC requirements and 4% FCM and dry matter intake generated from the prediction equations. Relationships between 4% FCM and nutrient concentrations were determined for heifers and multiparous cows. These recommendations suggest maximum feeding of nutrients to groups producing at least 35 kg of 4% FCM daily. The effect of grouping and individual feeding on 4% FCM were analyzed. Independent variables included nutrient concentration, dry matter intake, groups, calving interval, variability of 305 day milk production within the herd, and level of feeding. Holding dry matter intake and nutrient concentrations constant, increasing the number of groups from one to two increased 4% FCM by 0%-3%, two groups to three groups by 0-2%, and three groups to four groups by 0-1%. Changing from one group to individual feeding increased 4% FCM by 2%-4% and two group to individual feeding increased 4% FCM by 0%-3%. The range of figures was influenced by herd production level, calving interval, and variability of within herd production. High variability of production favored additional groups, while high calving intervals favored individual feeding. To estimate expected change in profitability due to a change in grouping or feeding methods, multiply previous milk revenue by expected percentage change and subtract the increased cost of feeding. Feeding less than the group average for nutrients was not found conducive for increasing profitability. In fact, production response to protein suggested that the low producing groups in multiple group feeding systems should be increased in protein slightly. Individual feeding had the most potential for profitability, although group feeding compared favorably under some circumstances. If cows were grouped, a minimum of two groups was generally preferable. If the high producing group did not satisfy the needs of the high producing cows, or the drop in nutrients between groups was substantial, a three group system would appear more favorable. / Ph. D.