The objective of this study was to determine the effects of nutrient allocation between the partial mixed ration (PMR) and pellet on feeding behavior and blood metabolites of mid-lactation dairy cows. A portion of an experimental diet was offered as a high-fiber (F; 7.7% starch, 41.2% NDF) or high-starch (S; 59.5% starch, 14.7% NDF) pellet fed at 1 or 3 kg twice per day. Four types of PMR were formulated for each pellet treatment to provide the same overall diet (PMR + pellet) among all treatments assuming DMI of 25.0 kg/d. Eight ruminally cannulated cows were used in a 4 × 4 Latin square design study with 14-d periods. Cows were fed PMR once daily at 1200 h, and pellet twice daily at 0600 and 1800 h. Disappearance of PMR was measured on d 11–13 of each period, as the amount of feed consumed every 3 h following PMR delivery, and blood was collected every 90 min on d 14. Treatment did not affect milk yield (42.7 kg/d), milk fat content (3.75%), or total dry matter intake (27.7 kg/d). In addition, there was no change in daily minimum, mean or maximum rumen pH, or evidence of PMR sorting among treatments. However PMR intake was reduced when more pellet was fed (22.9 vs. 25.3 kg/d; P < 0.01). When cows were fed the S pellet (with high-fiber PMR), PMR intake immediately after feed delivery was reduced as compared with the F pellet (28.5 vs. 33.5% of PMR intake; P = 0.04), and tended to be increased during other periods of the day. In addition, feeding the F pellet (with high-starch PMR) increased daily mean plasma concentrations of glucose (70.0 vs. 66.0 mg/dL; P < 0.01) and insulin (2.24 vs. 1.90 ng/mL ; P = 0.02), indicating that nutrient content of the PMR appeared to affect feeding behavior and blood metabolites to a greater extent than the pellet in the current study. These results suggest that nutrient allocation between PMR and pellet can affect actual PMR intake, feeding behavior, and plasma metabolites, and this should be considered when formulating diets consisting of PMR and pellet.