Bovine respiratory disease (BRD) is the main disease in beef production. Mass antibiotic treatment of high-risk calves at feedlot entry, a practice called metaphylaxis, is one of the most common tools to fight this disease. However, there is a growing tendency to minimize the use of antibiotics in food-producing animals.To develop a clinically based use of antibiotics, two experiments evaluated selective metaphylaxis with florfenicol (Nuflor®, Intervet-Schering-Plough S.A.;Boxmeer, Holland) administered to calves with a rectal temperature ≥ 39.7. °C (HT-calves).In the first experiment, 287 calves at high risk for BRD were randomly allocated to one of four experimental groups in a multi-site study HT-calves with florfenicol metaphylaxis (FHT), HT-calves without metaphylaxis (CHT), calves with a rectal temperature < 39.7. °C (LT-calves) with florfenicol metaphylaxis (FLT), and LT-calves without metaphylaxis (CLT). BRD incidence, average daily weight gain (ADG) and feed intake were recorded for 30 days. Respiratory symptoms were recorded in control calves to identify possible criteria to guide selective metaphylaxis.In the second experiment, 114 high-risk calves were randomly allocated to two groups in the FSM group, selective metaphylaxis with florfenicol. 33Nuflor®, Intervet Schering-Plough AH. was carried out in calves with a rectal temperature ≥ 39.7. °C (10/57 calves, 17.5%); in the TMM group, mass metaphylaxis with tulathromycine. 44Draxxin®, Pfizer AH. was carried out on all calves. BRD incidence, ADG, feed conversion, dressing percentage and lung lesions were monitored until slaughter. In the first experiment, we found no significant difference between FLT and CLT calves in BRD incidence (21.4% vs. 27.7%; P = 0.33) or ADG (1.16 vs. 1.21 kg/day; P = 0.46), or between FHT and CHT calves in BRD incidence (36.6% vs. 52.6; P = 0.24) or ADG (1.02 vs. 0.99; P = 0.83). Of all the clinical characteristics measured, only rectal temperature ≥ 39.7. °C was found to be significantly useful (P = 0.02) as a selection criterion for metaphylaxis.In the second experiment, no significant differences were found between FSM and TMM groups in BRD incidence (24.6 vs. 28.1%; P = 0.67), ADG (1.29 ± 0.17 vs. 1.34 ± 0.18 kg/day; P = 0.57), feed conversion (5.35 ± 0.36 vs. 5.34 ± 0.27 kg; P = 0.96), dressing percentage (48.61 vs. 49.30%; P = 0.19), or lung lesions (38.6 vs. 36.4%; P = 0.83). In addition, no BRD-associated mortality was observed, and only 1.69% of TMM calves developed chronic disease.Under our study conditions, selective metaphylaxis with florfenicol based on rectal temperature was associated with lower drug-related costs, less antibiotic use, and less intensive handling of calves than was mass metaphylaxis. © 2011 Elsevier B.V.