Metabolite identification in liquid chromatography−mass spectrometry (LC−MS) has several approaches. One of these approaches is using retention time (RT) with labeling technique. Using labeling technique, metabolite is easy to separate in LC-MS. Therefore, labeled metabolites get more clearly and discrete RT. Meanwhile, getting RT value take a lot of trouble and hard task. So, it is useful to predict RT with chemoinfomatic method instead of detecting RT. In this study, we predict RT of dansylated metabolite. Dansylation is one of the labeling techniques. It used at labeling metabolite containing amine or phenol. With this labeling technique, about 300 metabolites can be identified by LC-MS. But this is very robust and tired task because of RT. Too much RT appears, so it is hard to figure out what metabolite labeled is correspond with what RT. In this study, we use chemoinformatic method to predict RT of dansylated metabolites. In this method, we use MORDRED to calculate descriptors, Scikit-learn to select relevant descriptors, Artificial Neural Network (ANN) to build model. Moreover, we develop program using python and our RT prediction model to predict RT of dansylated metabolite. Also, we visualized our result used scatter plot. This scatter plot contains structure of dansylated metabolite. Moreover, when RT is shifted with some reasons like different experimental condition, our developing program can do RT correction and prediction. Specifically, when we know RT of some standard metabolite, we can make RT correction of these metabolites. Next, using this correction, we can predict RT of other metabolite. Therefore, we also predict RT of metabolite in different LC-MS conditions.