The objectives were to evaluate the effects of metaphylaxis (META) and pull-and-treat (PT) programs on health, antimicrobial use, beef production, economics, and greenhouse gas emissions in cattle at medium risk for bovine respiratory disease (BRD). A randomized complete block design was used at two US commercial feedlots. Steers and heifers [2366 total; 261 (±11.0) kg initial weight] were blocked by sex and feedlot arrival, and allocated to one of two pens within a block (16 pens total, eight blocks). Pens were randomly assigned to treatment: META, tulathromycin injection at initial processing; or PT, tulathromycin injection only for first clinical BRD treatment. Data were analyzed with linear and generalized linear mixed models. There was greater BRD morbidity in PT than META cattle (17.2% vs. 7.3% respectively; p < 0.01), and greater total mortality (2.5% vs. 1.1% respectively; p = 0.03). Per animal enrolled, 1.1 antimicrobial doses were used for META compared to 0.2 for PT (p < 0.01). Per animal enrolled, final live (p = 0.04) and carcass (p = 0.08) weights were greater for META than PT; however, net returns ($/animal) were not significantly different (p = 0.71). Compared to PT, total lifetime estimated CO₂ equivalent emissions from production were reduced by 2% per unit of live weight for META (p = 0.09). While antimicrobial use was reduced with PT, there may be substantial negative impacts on other outcomes if META was not used in this type of cattle population.