Our objective was to evaluate the effect of a health monitoring program based primarily on alerts generated by automated health monitoring (AHM) systems on disease detection and performance of dairy cows. Parous and nulliparous Holstein cows at 245 ± 3 d in gestation were stratified by parity and randomly assigned to a CON (n = 622) or TRT group (n = 621). Cows from both groups were commingled. For cows in CON, clinical examination (CE) was conducted daily up to 10 DIM and in response to daily milk yield reduction (≥15% negative deviation) alerts (Afimilk) or visual observation of clinical signs of disease from 11 to 30 DIM. For cows in TRT, CE up to 30 DIM was conducted in response to one or more of the following: an alert generated with a combination of rumination time and physical activity (HID score, SCR Dairy), milk yield deviation, or visual observation of clinical signs of disease. Daily after the morning milking, cows eligible for CE were evaluated following similar procedures. Binomial and quantitative data were analyzed by logistic regression and ANOVA with repeated measures, respectively. The proportion of cows with at least one event of metritis (P = 0.43; CON = 12.5%; TRT = 11.1%), displaced abomasum (P = 0.45; CON = 1.1%; TRT = 1.6%), indigestion (P = 0.73; CON = 2.9%; TRT = 3.2%), pneumonia (P = 0.3; CON = 1.5%; TRT = 0.8%), and mastitis (P = 0.4; CON = 10.3%; TRT = 8.9%) did not differ. The proportion of cows with ketosis (P = 0.09; CON = 8.7%; TRT = 6.1%) and the total proportion of cows with at least one event of disease (P = 0.05; CON = 30.4%; TRT = 25.3%) tended to differ. There was no difference in the combined proportion of cows sold and dead (P = 0.35; CON = 15.2%; TRT = 12.3%). No difference (P = 0.29) was observed for average weekly milk (CON = 41.6 ± 0.3 kg; TRT = 41.2 ± 0.3 kg) for up to 35 DIM. Cows inseminated at detected estrus (P = 0.30; CON = 66.7%; TRT = 63.8%) and pregnancy per AI to first service did not differ (P = 0.72; CON = 41.7%; TRT = 40.1%). We conclude that a health-monitoring program based primarily on alerts generated by AHM systems was an effective strategy to identify cows with health disorders and did not negatively affect herd production and reproductive performance outcomes. Supported by New York Farm Viability Institute project 017-014.