Twice weekly blood progesterone analysis was used to monitor resumption of ovarian cyclicity in 239 postpartum periods of 144 Holstein cows in one herd between D7 and D70 postpartum (PP). Calving season, low milk production and low BCS were risk factors for delayed cyclicity (DC: 16.3%). Retained placenta and metritis, early ovulation after calving and postpartum negative energy balance were identified as risk factors for prolonged luteal phase (PLP: 18.4%). At the same time, in 110 postpartum periods, ovaries were examined twice a week by ultrasonography. In normal cows (N: 56.5%), the first large follicle (>= 10mm) appeared 14.6 +/- 3.9 days after calving, ovulation occurred 28.9 +/- 9.5 days PP and the maximum diameter of the first corpus luteum (CL) was 24.9 +/- 4.9 mm. In DC cows, the interval between calving to the first large follicle tended to be longer than in normal cows (23.0 +/- 15.6 d; P=0.06). In PLP cows, the first large follicle was delayed (17.6 +/- 5.2 d, P<0.05), first ovulation occurred earlier (23.2 +/- 6.9 d PP, P<0.05), the maximum size of the first CL was not different from N cows (26.2 +/- 6.2 mm) despite its lifespan (33.6 +/- 13.8 days). Before first AI, cows not seen in oestrus or with a diagnosed abnormal progesterone profile were submitted to oestrus induction and synchronization treatment. The proportions of treated cows were 32.8, 62.9 and 60.5% in the N, DC and PLP groups of cows (significant difference between N and abnormal cows). In N cows, the interval from calving to conception was 117.3 +/- 60.6 days. Conception rate at first service (CR) was 31.3%, the incidences of early embryonic (EEM) and late embryonic-foetal mortality (LEM) were 47.9 and 33.3%. In DC cows, calving to conception interval was higher (144.0 +/- 67.3 d, P<0.05) than N cows. CR (28.6%) and EEM (60%) incidence were not different from N cows. In PLP cows, calving to conception interval was not different (132.8 +/- 62.0 d), CR was lower (16.3%, P=0.05) and LEM incidence tended lobe higher (56.3%. P<0.10) than in N cows. In conclusion, transrectal ultrasonography is not effective in distinguishing normal CL from CL associated with a prolonged luteal phase. PLP seems to have a negative effect on CR and on the incidence of LEM which should be investigated on a larger scale in Holstein cows.