Feed intake is frequently used in the dairy industry as an indicator of early lactation performance; however, it is difficult to assess individual dry matter intake (DMI) in group-housed cows. Therefore, measurable feeding behaviors in early lactation could be valuable in estimating DMI and predicting lactation outcomes. To investigate, feeding behavior collected during the first week of lactation from 5 studies at Kansas State University were compiled for analysis. Cows (n = 260) were housed in tie-stalls; feeding behaviors were recorded using individual feed bunks continuously measuring feed weight. Independent variables included meals (count/d), meal length (min/meal), and eating time (min/d). Additional data collected included wk 1 DMI (kg/day), parity, calving date, and the PTA for milk (PTAM). Variables were evaluated as either means or slopes over the 7-d period. Univariate analyses were used as a preliminary screen for predictors (P < 0.20) of DMI during wk 1 of lactation and 305MEM. Potential predictor variables were then included in a backward stepwise elimination model, along with associated quadratic and biologically relevant interaction terms, and the random effect of treatment nested within study. Criterion for model selection was minimal Bayesian Information Criterion. Feeding behavior in wk 1 was predictive of wk 1 DMI; meals/d, in addition to parity and PTAM (all P ≤ 0.006), were positively associated with wk 1 DMI (model r² = 0.29). Feeding behavior in wk 1 also predicted 305MEM; positive associations of PTAM (P < 0.001) and d 1 eating time (P = 0.009), along with negative associations of d 1 meals (P = 0.002) and wk 1 eating time (P = 0.01) combined to explain 50% of the variation in 305MEM. These results indicate that feeding behaviors in wk 1 of lactation can be used to predict DMI of individual cows as well as 305MEM. In addition, PTAM was retained in all models, indicating that genetic propensity for milk production is positively associated with early lactation DMI.