Determining the source of body fluids is crucial in forensic investigations, as it provides valuable information about suspects and the nature of the crime. Microbial markers that trace the source of tissues and body fluids based on site specificity and temporal stability are often used effectively for this purpose. In this study, a multiplex system comprising seven microbial markers (Finegoldia magna, Corynebacterium tuberculostearicum, Cutibacterium acnes, Haemophilus parainfluenzae, Streptococcus oralis, Prevotella melaninogenica and Faecalibacterium prausnitzii) was developed to distinguish between skin, saliva, and feces samples. Based on these markers, the system produces electropherograms that are specific for each sample type. We collected 492 samples from six different skin sites (palm, antecubital crease, inguinal crease, cheek, upper back, and toe web space), the buccal mucosa, and stool were collected to further test the system. Beta diversity analysis revealed distinct clustering among the three sample groups. Additionally, skin microenvironment cluster analysis was used to identify skin sites accurately. This analysis classified skin samples into four distinct microenvironments: dry, moist, oily, and foot. Finally, we established a machine learning prediction model based on random forest regression to identify the skin microenvironment, achieving an overall prediction accuracy of 79 %. The multiplex system developed in this study accurately identifies the sources of body fluids, and the skin microenvironment. These findings offer new insights into the application of microbial markers in forensic science.