Discrimination of food-contaminating microorganisms is an essential technology to secure the safety in manufacturing of foods and beverages. Conventionally, discrimination of the microorganisms has been performed by morphological observation, genetic analysis, and more recently, biochemical fingerprinting using mass spectrometry. However, several drawbacks exist in these methods, such as long assay time, cumbersome operations, and expensive equipment. To address these issues, we have proposed a novel method for discrimination of food-contaminating microorganisms, termed “colony fingerprinting”, based on bioimage informatics. In colony fingerprinting, growth of bacterial colonies were monitored using a lens-less imaging system. The characteristic images of colonies, referred to as colony fingerprints (CFPs), were obtained over time, and subsequently used to extract discriminative parameters. We demonstrated to discriminate 20 bacterial species by analyzing the extracted parameters with machine learning approaches, namely support vector machine and random forest. Colony fingerprinting is a promising method for rapid and easy discrimination of food-contaminating microorganisms.