As a critical post-translational modification, phosphorylation plays important roles in regulating various biological processes, while recent studies suggest that phosphorylation in bacteria is also critical for functional signaling transduction. Since identification of phosphorylation substrates and sites is fundamental for understanding the phosphorylation mediated regulatory mechanism, a number of studies have been contributed to this area. Since experimental identification of phosphorylation sites is time-consuming and labor-intensive, computational predictions attract much attention for its convenience to provide helpful information. However, although there are a large number of computational studies in eukaryotes, predictions in bacteria are still rare. In this study, we present a new predictor of cPhosBac to predict phosphorylation serine/threonine in bacteria proteins. The predictor is developed with CKSAAP algorithm, which was combined with motif length selection to optimize the prediction, which achieves promising performance. The online service of cPhosBac is available at: http://netalign.ustc.edu.cn/cphosbac/ .