The behavior of veterinary antibiotics in the soil is commonly studied using the following methods to add antibiotics to the soil: (A) adding manure collected from animals fed a diet that includes antibiotics; (B) adding antibiotic-free animal manure spiked with antibiotics; and (C) the direct addition of antibiotics. However, most studies have only used methods (B) and (C) in their research, and few studies have simultaneously compared the different antibiotic addition methods. This study used tylosin A (TYLA) as a model antibiotic to compare the effects of these three commonly used antibiotic addition methods on the dissipation rates of TYLA and the numbers of resistance genes in laboratory incubation experiments. The results showed that the three treatment methods produced similar TYLA degradation trends; however, there were significant differences (P<0.05) in the TYLA degradation half-life (t1/2) among the three methods. The half-life of TYLA degradation in treatments A, B and C was 2.44 ± 0.04, 1.21 ± 0.03 and 5.13 ± 0.11 days, respectively. The presence of manure resulted in a higher electrical conductivity (EC), higher relative abundance of Citrobacter amalonaticus, higher macrolide resistant gene (ermB, ermF and ermT) count and lower ecological toxicity in the soil, which could partially explain the higher TYLA degradation rate in the treatments containing manure. The higher degradation rate of TYLA in treatment B when compared to treatment A could be due to the lower concentrations of tylosin B (TYLB) and tylosin D (TYLD). The main route for veterinary antibiotics to enter the soil is via the manure of animals that have been administered antibiotics. Therefore, the more appropriate method to study the degradation and ecotoxicity of antibiotic residues in the soil is by using manure from animals fed/administered the particular antibiotic rather than by adding the antibiotic directly to the soil.