Six Holstein dry cows were used to evaluate the effect of a sensory additive (ProEfficient, Lucta S.A.) on eating behavior in a complete randomized design. Cows divided into 2 pens were randomly assigned to 2 treatments: control TMR or the same TMR supplemented with ProEfficient (PE) at a dose of 15 g/cow/d. The TMR (73:27 concentrate: forage; 14.8% CP, 47.0% NDF, 1.32 Mcal NEL/kg) was formulated to provide the nutrient requirements of dry cows following NRC (2001) recommendations. Cows were fed ad libitum through 4 automatic feeders mounted on scales within each pen. Feed consumption and feeding bouts were recorded automatically. Data were analyzed with a mixed-effects model with repeated measures using the PROC MIXED procedure of SAS (1999) with cow as a random effect and treatment, time, and their 2-way interaction as fixed effects. Although DM intake was not affected (P > 0.05) by treatments (11.4 kg/d ± 0.33 SE), cows receiving PE reduced (P < 0.05) the time dedicated to eat (119 vs. 149 min/d± SE 7.60) whereas increased (P < 0.05) the eating rate (106.4 vs. 85.3 g/min ± 5.52 SE) compared with the control cows. Neither number of meals (4.3 ± 0.75 SE) nor meal size (2.66 kg/meal ± 0.19 SE) differed (P > 0.05) between treatments. Feeding a sensory additive increased eating rate without affecting total dry matter intake of dry cows.