A fast and fully automated image analysis technique for the enumeration of fluorescence-labeled bacteria in oral sampleswas developed. This paper describes the system configuration, application strategy, automated operation, and initial validation experiments using fluorescent microspheres, bacterial cultures, in vitro grown biofilms and human dental plaque. Following a series of brief operator-controlled calibration steps, the technique automatically performs all necessary microscope operations (stage translation, focus, sampling and analysis) on slides with up to 48 wells for as many different samples. It quantifies bacteria from differential interference contrast images, images showing cells that had been labeled by immunofluorescence with monoclonal antibodies, or images with cells labeled by a fluorescent DNA stain. With all evaluated samples, close agreement between the automated system and the assessor's visual counts was observed. This novel automated image grabbing and analysis procedure is applicable to the enumeration of specific taxa in clinical samples by both immunofluorescence and fluorescent in situ hybridization.