Milk fat production is highly influenced by nutrition and rumen fermentation. Rumination is an essential part of the ruminant digestive process and can serve as an indicator of rumen fermentation. The objective of this research was to quantify variation in rumination time between and within dairy herds and to determine relationships between rumination time and milk fat production and fatty acid (FA) profile as a proxy of rumen fermentation. Our hypothesis was that rumination may indicate disruptions to rumen fermentation and that lower-rumination cows would have lower milk fat due to these rumen disruptions. Data were collected from 1733 Holstein cows on 5 commercial dairy farms (4 in PA and 1 in NY) of 200 to 700 head using one of 2 commercially-available rumination sensing systems (RSS). Rumination data were collected for 7 consecutive days leading up to a DHIA test, summed within day, then averaged to obtain average daily minutes of rumination time. Milk samples from the DHIA test were analyzed for fat content by mid infrared spectrum and for milk FA profile by gas chromatography. Rumination data were analyzed using multiple linear regression models in JMP Pro v13.0 and R v3.4.4 (SAS Institute, Cary NC; R Core Team, Vienna, Austria). Rumination time was related to concentration of specific odd and branched-chain and trans-FA in milk but was not directly related to milk fat concentration. Rumination also did not contribute to models predicting milk fat concentration. The relationship between trans-10 C18:1 and rumination was significant (P < 0.001) and positive after accounting for the effect of farm, with a partial R² of 2.97%, 4.24%, and 2.22% across all data, RSS 1, and RSS 2 data respectively. While rumination was not related directly to milk fat, it was associated with differences in trans and odd and branched-chain FA that also change during subacute ruminal acidosis or milk fat depression, which may impact milk fat and other production variables. Further investigation using rumination data from commercial systems to predict or identify the presence of these conditions is warranted.