Since the prohibition of antibiotics as animal growth promoters, demand for effective probiotic strains has steadily increased. The goal is to maintain productivity and mitigate environmental concerns in the livestock industry. There are many probiotic animal-diet supplements available, over 2,000 products in the Republic of Korea alone, with little explanation about the desirable properties of each probiotic strain. The purpose of this study was to describe the underlying logic and methods used to select two novel strains of probiotic candidates. To economically screen these candidates, the abundance of surfactin secreted was used as an in vitro marker. We used a modified oil-misting method to screen ~2,000 spore-forming bacteria for novel strains of Bacillus subtilis. Of these, 18 strains were initially selected based on the semiquantitative criterion that they secreted more surfactin than B. subtilis ATCC21322 on Luria-Berani (LB) agar plates. The whole genome sequence was determined for two of the 18 strains to verify their identity. A phylogeny of 1,162 orthologous genes, genome contents, and genome organization confirmed them as novel strains. The surfactin profiles produced by these two strains consisted of at least four isoforms similar to standard surfactin and enhanced cellulase activities up to 50%. Four fractionated individual isoforms of surfactin suppressed inflammation induced by lipopolysaccharides. The half-maximal inhibitory concentration (IC50) was about 20 μM for each isoform. Both selected strains were susceptible to seven important antibiotics. Our results implied that an abundant secretion of surfactin was a useful biomarker in vitro and could be utilized for mining probiotic candidates through high-throughput screening of environmental samples.