The objectives of this study were to quantify the effect of an intrauterine infusion of cephapirin on the reproductive performance at first service of postpartum dairy cows affected by purulent vaginal discharge (PVD) or cytological endometritis (ENDO) using different diagnostic strategies and to determine if the presence of prolonged anovulation would influence the magnitude of treatment benefit. A total of 2,259 Holstein cows in 28 herds were enrolled in a randomized clinical trial. At 35 (±7) DIM, cows were diagnosed for PVD (purulent vaginal discharge or worse using the metricheck device) and ENDO (≥6% polymorphonuclear cells using the cytobrush technique or at least small amounts of leukocytes using the leukocyte esterase colorimetric test). Regardless of reproductive tract disease status, cows were randomly assigned to receive an intrauterine cephapirin infusion or no treatment. Serum progesterone was measured at 35 and 49 (±7) DIM (14 d apart); cows were considered to have prolonged anovulation if progesterone was <1 ng/mL at both times. Statistical analyses were conducted using multivariable mixed logistic regression models adjusted for confounders and herd clustering effect. Intrauterine cephapirin treatment was associated with an increased first service pregnancy risk in cows diagnosed with PVD (no treatment: 15.4%; treatment: 31.4%; P < 0.05) and ENDO (cytobrush: no treatment: 16.2%, treatment: 24.4%, P < 0.05; leukocyte esterase: no treatment: 15.8%; treatment: 25.1%, P < 0.05), but not in cows unaffected by any form of reproductive tract disease (no treatment: 34.8%; treatment: 32.6%; P = 0.5). The effect of cephapirin treatment in anovular cows (no treatment: 21.0%; treatment: 26.4%; P = 0.26) was numerically lower than in cyclic cows (no treatment: 22.7%; treatment: 34.1%; P < 0.05). Overall, an intrauterine infusion of cephapirin improved first service pregnancy risk in cows with postpartum reproductive tract disease and this effect was influenced by postpartum anovulation status.