This study evaluated the association between estrous expression and the duration of the current and previous follicular and luteal phases. A total of 169 events from Holstein cows were enrolled. Cows were monitored continuously by an accelerometer. Ovaries were scanned from the day of first alert until pregnancy as follows: day of alert (0 d), 24 h, 48 h, 10, 14, 17, 21, 24, 28, 31, 35, 38 and 42 d post-alert. Blood samples for progesterone (P4) analysis were collected at the same times as described above. Area under the curve (AUC) was calculated for P4 concentration over the total cycle. Duration of the follicular phase (FPh) was considered as the day of luteolysis until next ovulation and luteal phase (LPh) as the time of ovulation until next luteolysis. Luteolysis was considered as 2 consecutive small CLs (<25mm). Estrous expression was classified as high (≥80 index) and low (<80 index) using the median. The LPh was categorized as short (<18 d) or long (≥18 d) also using the median. Data were analyzed by the GLIMMIX procedure of SAS with estrus event as experimental unit and cow as a random effect. An average of 4.4 ± 2.1 estrous cycles per cow were recorded. Each unit increase of AUC of P4 was associated with a 0.15-d increase in the LPh (P < 0.01). The duration of the preceding LPh had an impact on subsequent FPh, as an increase of 1 d of the LPh was associated with a 0.30 d increase in the FPh. Cows with high estrous expression had shorter LPh, FPh and overall length of the estrous cycle when compared with cows that had low estrous expression (21.3 ± 1.1 vs. 25.6 ± 1.9d, P = 0.06; 6.5 ± 0.5 vs. 8.4 ± 0.9 d, P = 0.10; and 27.7 ± 1.7 vs. 37.5 ± 3.1 d, P = 0.02, respectively). Higher concentrations of P4 at the time of estrus were associated with a shorter LPh (slope = −0.10; P < 0.01) but not FPh. Shorter LPh before AI had higher pregnancy per AI (P/AI) when compared with cows with a longer LPh (50.5 ± 7.0 vs. 27.9 ± 8.0%, respectively; P = 0.03). Additionally, for every day increase in FPh duration there was a tendency for 2.5% increase in P/AI (P = 0.10). In conclusion, the duration of the luteal phase is affected by P4 during the preceding cycle and at the time of estrus, as well as by the intensity of estrous expression detected by an automated activity monitor.