Our objective was to predict dry matter intake (DMI) by lactating cows using factors related to the filling effect of rations. A database of 156 treatment means from 44 experiments reported in the literature was developed. The database included data for cows ranging from 60 to 309 d postpartum and included diet forage NDF (FNDF) content (23.6 ± 5.8, mean ± SD), diet ADF/NDF (0.61 ± 0.08, mean ± SD), and a laboratory measure (in vitro or in situ) of NDF digestibility (LNDFD) of the sole forage or major forage (51.2 ± 12.1, mean ± SD). Models included the random effect of study as well as its interaction with LNDFD to account for differences in methods used to determine LNDFD. The full model also included linear and quadratic effects of FNDF, ADF/NDF, and LNDFD, as well as their linear and quadratic interactions, mean milk yield (MY) for each study and its interaction with the diet factors. Equations were developed by stepwise regression with backward elimination and treatment means were weighted by the inverse of their variance. The prediction equation for DMI (r² = 0.80, RMSE = 2.76 kg/d) is: DMI (kg/d) = 13.4 - 0.078 x FNDF + 8.264 x ADF/NDF + 0.0126 x LNDFD + 8.453 x (ADF/NDF-0.602)² + 0.178 x MY – 0.172 x (ADF/NDF-0.602) x (MY-33.1) + 0.0060 x (LNDFD-48.2) x (MY-33.1) + 8.183 x (ADF/NDF-0.602)² x (MY-33.1). DMI was positively related to MY and ADF/NDF and negatively related to FNDF, while LNDFD was negatively related to DMI for cows with low MY but positively for cows with high MY. Response to higher ADF/NDF was greater as MY increased. ADF/NDF was included to represent differences in forage fragility between grasses and legumes but it is also affected by the fraction of cereal grain in the diet. The following equation was developed for when LNDFD data is not available (r² = 0.77, RMSE = 3.01 kg/d) using the same database: DMI = 11.6 – 0.097 x FNDF + 8.31 x ADF/NDF + 0.268 x MY. DMI was related positively to MY and ADF/NDF and negatively to FNDF with no interactions detected with MY. These equations might be useful to evaluate DMI response to factors related to the filling effects of rations.