Individual energy balances (EB) and total net energy intake (NEI) were modelled to quantify the energy content of the explanatory variables body-weight change corrected for gut fill (BWC) and body condition-score change (BCSC). A total of 6147 records from 213 lactating dairy cows, 71 Holstein-Friesians, 71 Jerseys and 71 HolsteinJersey F1-crosses was analysed (1st, 2nd and >2nd lactation). Data were collected during 210 days of lactation, from calving to week 30. Individual EB were calculated as total energy for lactation (MJ NEL) minus estimated energy necessary for maintenance and production. Body-weight changes were calculated as differences between weekly means of body weight corrected for gut fill (BWCw) as well as from polynomial regression (BWCp). BCSC were differences between assessed body condition scores. The study included analyses of 1- and of 4-week periods for BWCw and BWCp, of differences between measuring periods for BCSC as well as separate analyses of data from cows in either negative or positive energy balance.

Analysis of repeated measurements were performed applying mixed model theory. Models for EB contained type of dairy cow, lactation and type of roughage as fixed effects, week of lactation and either BWCw, BWCp or BCSC. Models for NEI contained type of dairy cow, lactation and quality of roughage as fixed effects, week of lactation as well as energy corrected milk, metabolic body weight and either BWCw, BWCp or BCSC as covariates.

Most plausible energy contents per unit body tissue change were estimated for BWCp from weekly data with either EB (with energy partitioned into milk production and maintenance fixed) or NEI (no restrictions concerning partitioning) as the response variable. The estimated energy contents for BWCp were 15·40 and 10·47 MJ NEL, respectively.