In a commercial dairy herd, 316 lactating Holsteins were studied to determine the percentage of anovular cows, to examine follicular sizes in anovular cows, and to compare synchronized ovulation (Ovsynch) versus detection of estrus on fertility of ovular and anovular cows. Ultrasonography examinations at 47 to 53 d and at 54 to 60 d postpartum were used to measure follicles and to classify cows as ovular or anovular. Anovular cows were identified as those with no detectable luteal tissue by ultrasonography and by low progesterone in blood samples collected weekly. Anovular cows included 28% of 122 primiparous cows and 15% of 194 multiparous cows. Of 64 anovular cows, 20% had follicles > or = 25 mm that might be considered cystic (4% of total cows), 58% had 15- to 24-mm follicles, and 22% had 9- to 14-mm follicles. Cows identified as ovular and anovular were randomly assigned within cyclic status to one of two artificial insemination (AI) strategies: 1) AI after detected estrus during 21 d, or 2) timed AI after a 10-d Ovsynch protocol. Weekly ultrasonography continued for 21 d to detect ovulations. For the Ovsynch sub-groups, 97% of ovular and 94% of anovular cows ovulated after the second GnRH injection. Within 21 d, spontaneous ovulations for the detection of estrus sub-groups were 42% of anovular cows vs. 89% of ovular cows. Conception rates were greater for ovular cows regardless of treatment, but conception rates between respective Ovsynch and estrus detection groups for ovular (32%, 35%) or anovular (9%, 11%) cows were similar. Although 20% of lactating cows were not cyclic by about 60 d postpartum, nearly all ovulated following Ovsynch. However, anovular cows had lower conception than ovular cows whether inseminated after detected estrous or after Ovsynch.