The objectives of the study were to validate the use of endometrial cytology (EC) and ultrasonography (US) to diagnose subclinical endometritis in clinically normal postpartum dairy cows, and to measure the impact of subclinical endometritis on reproductive performance. Holstein cows from two dairy farms were examined at Visit 1 (V1) at 20-33 days in milk (DIM), and clinically normal cows (n = 228), based on the absence of abnormal discharge on external inspection and vaginoscopy, were selected. The reproductive tract of selected cows was evaluated by transrectal palpation, US and EC. All cows in the study were re-examined at Visit 2 (V2) at 34-47 DIM (2 weeks after V1) and were subsequently followed for a minimum of 8 months (until pregnant or culled). Survival analysis was used to derive a case definition of subclinical endometritis, based on factors associated with decreased relative pregnancy rate. Positive EC at V1 (>18% polymorphonuclear leukocytes; PMN) or fluid in uterus at V1 (FIU1) were associated with a significant reduction in the relative pregnancy rate and identified cows with subclinical endometritis. Similarly, a positive EC (>10% PMN) at V2 or fluid in the uterus at V2 (FIU2), identified cows with subclinical endometritis. Cows with subclinical endometritis at V1 and at V2 had a relative pregnancy rate of 41 and 51% (hazard ratio for pregnancy of 0.59 and 0.49), respectively, compared to cows without subclinical endometritis. Given EC or US findings, no diagnostic criteria based on transrectal palpation of the uterus had predictive value for risk of pregnancy. In conclusion, subclinical endometritis, diagnosed by EC or US, was associated with reduced relative pregnancy rate.