IntroductionThe aim of this study was the evaluation of nutrient excretion patterns in samples of feces collected every 2 h to define the best sampling protocol for estimation of apparent digestibility.MethodsFour multiparous mid-lactation Holstein cows, housed in a tie stall barn and milked twice a day (0800; 1900 h), were enrolled. Dry total mixed ration (TMR) without silages was fed once (0800 h) per day. Feces were sampled every 2 h for 72 h. Each sample was divided in 3 portions: hourly sample sample (8, 10, 12, 14, 16, 18, 20, 22, 00, 2, 4, 6), 8-h composite sample (00–06, 08–14 and 16–22), and a 24-h composite sample. Complete chemical analyses were performed and total tract nutrient digestibility was calculated using undegraded neutral detergent fiber at 240 h of in vitro fermentation (uNDF240h) as a marker. Feeding and rumination patterns were also recorded during the trial.Results and discussionFor some parameters, excretion was not constant throughout the day: neutral detergent fiber (aNDFom), undegraded neutral detergent fiber at 24 h of in vitro fermentation (uNDF24h), uDNF240h, total tract crude protein digestibility (TTCPD), total tract neutral detergent fiber digestibility (TTaNDFomD), total tract potentially degraded neutral detergent fiber at 240 h of in vitro fermentation digestibility (TTpdNDF240hD) with minimal values after new TMR delivery and maximal values 12 h after feed delivery. Feeding and ruminating behavior seemed to have an important role in the excretion pattern, due to the pushing and evacuating effect they have. Considering our results, two fecal samples at 12 and 24 h after the TMR delivery are suggested. For one daily sample, 12 h post time of most stable and constant rumination 0000–0600 h, which is also 8 h post feed delivery is suggested.