To assess the effect of feeding strategy (pasture vs. total mixed ration, TMR) on diurnal feeding pattern a study was conducted in a randomized block design for 20 d (15 d of adaptation and 5 d of measurements). Fourteen Holstein dairy cows (DIM = 148 ± 4.7; BW = 535 ± 10.9 kg; BCS = 2.6 ± 0.08; milk yield = 28.9 ± 0.9 kg) were allowed to graze an oat pasture (herbage allowance = 45 kgDM/day; 165 d from seeding, DM = 38% at ground level, NEL = 1.56 Mcal/kgDM) and were supplemented at milking parlor (0.9% BW, corn and barley grain as energy source and soybean meal as protein source), PAS treatment, or were confined and individually fed ad libitum TMR based on corn silage (55:45 forage to concentrate ratio, as dry basis; DM = 40%, NEL = 1.73 Mcal/kgDM), offered once daily in the morning (TMR treatment). Cows were milked twice daily and the access time to each diet was 16 h. Individual cows were observed during light hours every 5 min on 3 consecutive days and time spent eating, ruminating or other (idling and drinking) was recorded assuming that the activity was maintained during each 5 min. An event was defined when the activity was maintained for at least 2 consecutive observations. Data were analyzed using a mixed model and repeated measures, with treatment as fixed effect and block and day as random effects. The time spent eating was greater while time spent in other activity was lower (P < 0.001) for PAS than TMR cows (283 vs. 227 ± 14 min for grazing and 110 vs. 183 ± 12 min for other, respectively). Duration of the first meal events after the am and pm milking did not differ between diet types and averaged 69 ± 8 min and 43 min ± 7, respectively. However, the first non-eating event after the am milking (85 vs. 55 ± 13 min) and the last meal event before pm milking (43 vs. 27 ± 8 min) were greater (P < 0.01) for PAS than TMR cows. Results suggest that different ingestive behavior performed by TMR and PAS cows could be associated with different dietary signals that would affect short-term feed intake.