The objective of this study was to compare the cash flow of lactating Holstein cows randomized to health monitoring strategies that relied exclusively on automated health alerts or visual observation for selecting cows for clinical examination from 3 to 21 days in milk (DIM). Data from a randomized controlled trial was used to estimate total cash flow per cow and per slot up to 100 DIM. We expected that greater milk yield in early lactation for the group monitored with automation would result in greater cash flow than for the group including visual observation despite greater health monitoring and treatment costs. Lactating Holstein cows (n = 1,192) fitted with a neck-attached automated rumination time and physical activity monitoring system (Sensehub, Merck Animal Intelligence) and that had milk weights collected automatically at every milking (MM27BC, DeLaval) were randomly assigned to the visual observation (VO = 594) or automated health monitoring (AHM = 598) treatment. Cows were selected for clinical examination based on automated health alerts (i.e., health index score <86 arbitrary units, daily rumination time <250 min, and reduction in daily milk yield >20%) in the AHM treatment and exclusively based on visual observation of clinical signs of disease in the VO treatment. Health, milk yield, and herd exit outcomes, and prices for inputs and outputs were collected through 100 DIM. Cash flow per cow and per slot (unit of space occupied by a cow at a dairy) including income over feed costs, health monitoring and management costs, treatment costs, and replacement costs were estimated and compared using deterministic analysis with linear models and stochastic analysis. Regardless of the method of estimation for cows that remained in the herd, cows that exited the herd, or for all cows combined, cash flow differences from the deterministic analysis generally favored the AHM treatment. Likewise, the stochastic simulations supported the hypothesis that most of the time, the AHM treatment would result in positive cash flow compared with the VO treatment. Notably, the magnitude of the differences between treatment groups varied substantially based on the effects of treatments on herd performance outcomes, the herd exit dynamics, prices for influential inputs and outputs, and the methods used to estimate cash flow. Implementing a fresh cow health monitoring program that relied exclusively on automated monitoring systems alerts to select cows for clinical examinations was economically beneficial compared with a program that relied exclusively on visual observation to select cows for clinical examinations.