Multiparous ruminally cannulated Holstein cows (n = 16) were fed either a lower starch [21% starch, 37% neutral detergent fiber (NDF)] or higher starch (27% starch, 32% NDF) diet from 1 to 21 d in milk (DIM) to assess the effect of dietary starch on rumen fermentation, rumination, energy metabolites, and inflammatory markers. Diets contained 28% corn silage, 22% haycrop silage, 2% straw, and 48% concentrate. Corn meal was replaced partially with soyhulls and wheat middlings in the lower starch diet. Cows were fed a close-up diet (16% starch, 44% NDF) for 21 d. Cows were housed in a bedded pack, fed individually, and milked 3x/d. Rumen pH was measured at 1-min intervals and rumination was measured continuously from −14 to 21 DIM. Rumen and blood samples were collected on −14, 1, 2, 3, 4, 5, 6, 7, 9, 13, 17, and 21 DIM at 0 and 6 h post feeding. Data were analyzed as a completely randomized design by ANOVA with the MIXED procedure of SAS with model effects of treatment, time, treatment x time, and covariate (data from −14 to −8 DIM). Cows averaged 20.4 ± 0.7 kg dry matter intake/d and 42.1 ± 1.9 kg milk/d with 4.6 ± 0.2% fat and 3.5 ± 0.1% true protein. Daily mean rumen pH was higher (6.19 vs. 6.04, SE = 0.05, P = 0.03), time with pH <5.8 was lower (1.6 vs. 4.5 h/d, SE = 0.7, P < 0.01), and rumen ammonia was higher (10.1 vs. 7.1 mg/dL, SE = 0.6, P < 0.01) for cows fed the lower starch diet. Diet did not affect (P > 0.10) rumination (481 ± 18 min/d), rumen volatile fatty acids (115 ± 3 mM), or acetate to propionate ratio (3.23 ± 0.13). Cows fed the lower starch diet had lower (P < 0.10) 6-h serum haptoglobin [0.10 (95% CI 0.06–0.14) vs. 0.17 (95% CI 0.11–0.27) mg/mL] and serum amyloid A [22 (95% CI 12–38) vs. 53 (95% CI 28–101) ug/mL]. Diet did not affect (P > 0.10) 0-h nonesterified fatty acids (680 ± 58 uEq/L), 0 and 6-h β-hydroxybutyrate (1.0 ± 0.2 mM), 6-h interleukin (IL) −1β (57.3 ± 5.1 pg/mL), 6-h IL-6 (457 ± 42 pg/mL), or 6-h tumor necrosis factor-α (40.7 ± 2.7 pg/mL). The nutritional strategy that is used during the transition period is critical for minimizing the risk of SARA and controlling inflammation.