- The objective of this study was to investigate, describe, and quantify daily body weight (BW) changes in the first 120 d of lactation in high-producing dairy cows. Data included 255,287 daily BW measurements from 2,167 Israeli Holstein dairy cows originating from 7 commercial dairy farms. Individual series of measurements were first smoothed using cubic splines for generating variables representing BW changes in early lactation and further analysis of the data. To construct standard BW curves stratified by parity and adjusted for farm, mixed models for repeated measurements were fit to the smoothed data, and least squares means for day in lactation were plotted. Time-series analysis techniques using polynomial functions of day in lactation and pairs of sine and cosine functions representing 7- and 21-d cycles were performed separately on each individual series of measurements. Additionally, generalized estimating equations were used to perform similar analysis on the data set as a whole. Mean days from calving to nadir BW increased significantly from first to later parities, as did mean BW loss from calving to nadir. The first-parity cow lost 6.5% of her BW from calving to d 29 in lactation, and second-parity and greater-parity cows lost 8.5 and 8.4% of their BW to d 34 and 38 in lactation, respectively. After nadir BW was reached, first-parity cows regained relative BW at a greater rate than did older parity cows. The trend in BW was nonlinear. A 7-d cycle was present in 247 cows (11.4%) and a 21-d cycle was present in 715 cows (33.0%). Presence of a 21-d cycle was associated with a 33% reduction in the risk of being diagnosed with inactive ovaries. Fewer days from calving to nadir BW and smaller BW loss from calving to nadir, coupled with a faster post-nadir increase in relative BW in first-parity cows compared with older cows indicated a smaller energy deficit in early lactation. Association between 21-d cycles in BW and ovarian activity suggest that these cycles were physiological and related to the estrous cycle. Therefore, monitoring them could be useful for indirectly assessing ovarian activity in a herd.