Serum total protein fluctuation, during the first 70 days of age, under different milk feeding programs, were evaluated in thirty-seven calves, males and females, of Holstein breed (15/16 Holstein: Guzerá, high grade and pure breed) . Calves at birth were alloted to one of the following treatments: A1 - Whole milk, colostrum and milk replacer (1:1:2) during the first 30 days of age; A2 - whole milk, colostrum and milk replacer (1 :1 :2) during the first 45 days of age; A3 - whole milk, colostrum and milk replacer (1:1:2) during the first 60 days of age; B - whole milk during the first 45 days of age and C -milk replacer during the first 45 days of age. All calves were fed colostrum up to 72 hours after birth. The employed commercial milk re placer had milk sub-product as protein source. The concentrate used was a commercial product for early weanning. The blood samples were collected in the following times after birth: zero (before suckling), 1, 2, 5, 10, 15, 20, 25, 30, 40, 50, 60 and 70 days post-partum. Serum total protein was analysed by the biuret method. The results were interpreted by the following statistical analysis: analysis of variance, Tukey, correlations, regressions and analysis of variance under factorial model considering the effects of treatments, sampling time and interactions. The lowest concentration of serum total protein was found at birth (5,02 ± 0,51 g/100 ml) and the highest (8,09 ± 0,51 g/100 ml) at 24 hours post-partum. In the first 30 days, the concentration in the treatment A (A1 + A2 + A3) was significantly higher than treatment C and not different from B. It was found a linear and quadratic reduction of total protein concentration, in the first 30 days, followed by an increase from 50 to 70 days.