The objective of this study was to determine risk factors related to increase in physical activity during estrus. Estrus events (n = 2,510) from 1,358 lactating Holstein cows from 2 commercial dairies were recorded. Cows were monitored continuously by 3 automated activity monitors (AAM). A collar-mounted accelerometer (Collar; Heatime, SCR Engineers) and a leg-mounted pedometer (Leg1; Boumatic Heat-seeker-TX) were used in farm 1. A second leg-mounted pedometer (Leg2; AfiMilk Pedometer Plus) was used in farm 2. For cows with collar and Leg1, only spontaneous estrus (SE) events were registered (1,099 events; 318 cows). Cows with Leg2-tags were induced to ovulate (IO; 1,411 events; 1,040 cows) through an E2/P4-based TAI protocol. Body condition score (BCS; 1 to 5 scale) was measured at moment of AI. SE cows’ also had secondary signs of estrus (clear vaginal mucus, uterine tonus, signs of mounting) recorded. Milk production was recorded at each milking. Ovarian ultrasonography was performed at estrus. Data were analyzed by Pearson correlation and ANOVA. Collar, Leg1 and Leg2 systems’ peak of activity (PA) and Collar’s duration (DU) were affected by parity as multiparous expressed lower PA (69.3 ± 0.8 vs. 75.9 ± 1.1 index [Collar]; 323.9 ± 6.0 vs. 354.8 ± 8.5% [Leg1]; 323.04 ± 4.6 vs. 336.3 ± 5.3% [Leg2]) and shorter duration (10.7 ± 0.2 h vs. 12.0 ± 0.3 h) than primiparous. Cows with BCS ≤2.5 showed less intense estrus measurements (PA and DU). Lactation stage did not influence PA or DU of estrus. Cows that displayed secondary signs of estrus had greater PA and DU measured by Collar and Leg1 systems. Intensity progressively increased as more secondary signs were displayed. Milk production was negatively correlated with PA (r = −0.20), but no difference was observed in PA and DU when milk production was categorized into quartiles. Quantitative data from activity monitors can be used to survey estrus behavior in dairy cows, but must be calibrated based on risk factors to ensure accurate behavior interpretation.