The objective was to investigate the economic effect of treating dairy cows with metritis using ceftiofur-free acid or leaving them untreated at the time of diagnosis. Cows with a fetid, watery, red-brownish vaginal discharge were diagnosed with metritis (d 0). Data from 875 dairy cows (506 primiparous and 369 multiparous) from 1 herd in northern Florida that had been part of a larger study evaluating different treatments for metritis were used for the economic analysis. Holstein cows with metritis had been randomly assigned to: Ceftiofur (CEF, n = 239) = subcutaneous injection of 6.6 mg/kg of ceftiofur crystalline-free acid in the base of the ear at d 0 and d 3; Untreated (UNT, n = 233) = no treatment applied at metritis diagnosis. Both groups could receive escape therapy if condition worsened. A group of nonmetritic healthy cows (NMET; n = 403) from the same cohort was randomly selected for comparison. Continuous outcomes such as 300-d milk production (kg/cow), milk sales ($/cow), cow sales ($/cow), treatment cost by 60 days in milk ($/cow), reproduction cost ($/cow), replacement cost ($/cow), feeding cost ($/cow), and gross profit per cow ($/cow) were analyzed using the ANOVA (MIXED procedure of SAS version 9.4). Dichotomous outcomes such as pregnancy and culling by 300 d were analyzed using logistic regression (GLIMMIX procedure of SAS). Models included the fixed effects of treatment, parity, and the interaction between treatment and parity. A stochastic analysis was performed with 10,000 iterations using the observed results from each group. The CEF treatment resulted in greater treatment cost by 60 DIM than UNT ($112 vs. $37), but resulted in a greater proportion of pregnant cows (71 vs. 61%) and decreased culling by 300 DIM (29 vs. 39%) compared with UNT. Gross profit was lesser for UNT than NMET ($2,969 vs. $3,426), and CEF was intermediate ($3,219). The stochastic analysis showed that the mean difference in gross profit between UNT and NMET was -$457; saleable milk (49%) and replacement cost (24%) accounted for most of the variation. The mean difference in gross profit between CEF and NMET group was -$207; saleable milk (82%) and initial metritis treatment cost (9%) accounted for most of the variation. The mean difference in gross profit between the UNT and the CEF group was -$250; replacement cost (41%) and cow sales (31%) accounted for most of the variation. In summary, metritis caused large economic losses when left untreated, and CEF reduced those losses by improving fertility, reducing culling and replacement cost, and reducing milk yield losses.