The objective of the study was to evaluate if milk urea nitrogen (MUN) and milk fat content could be used as the noninvasive indicator to identify cows with a greater or lower risk of subacute ruminal acidosis (SARA). Our hypothesis was that cows greatly very in their risk to develop SARA even if they are fed a common diet, and that cows with lower MUN and milk fat content would have a greater risk of SARA while cows with higher MUN and milk fat content would have a lower risk of SARA. In a screening study, 35 late-lactating Holstein cows (DIM = 250 ± 71.1; BW = 601 ± 45.4 kg) were fed a high grain diet containing 35% forage and 65% concentrate mix ad libitum for 21 d. Concentration of MUN ranged from 5.7 to 13.9 mg/dL among the 35 cows, and the average milk fat content was 3.5%. Then, 5 cows with highest MUN concentrations with milk fat higher than 3.5% were selected as animals that presumably have a low risk of SARA (LOW), and 5 cows with lowest MUN concentrations with milk fat less than 3.5% were selected as animals that presumably have a high risk of SARA (HIGH). Then these 10 animals were ruminally cannulated during the subsequent dry period. As one LOW cow was culled due to fatty liver, 9 animals (DIM = 122 ± 33.2; BW = 615 ± 49.1 kg) were used for the subsequent study in the following lactation. All cows were fed a high grain diet consisting of 35% forage and 65% concentrate mix ad libitum for 21 d. Ruminal pH was measured every 30 s for 72 h. All response variables were evaluated for the group effect using the PROC TTEST procedure of SAS (version 9.2, SAS Institute Inc., Cary, NC). Minimum (5.75 vs. 5.30; P = 0.02) and mean ruminal pH (6.35 vs. 6.04; P = 0.02) was higher for LOW compared with HIGH animals. In addition, duration of rumen pH below 5.8 was shorter in LOW animals (52.5 vs. 395 min/d; P = 0.04). These results suggested that MUN and milk fat content in late-lactating cows fed a high grain diet may be used to identify cows that have a higher or lower risk of SARA.