The microbiome is a promising tool for identifying body fluids which can be deposited on various substrates at a crime scene. Body fluids collected from crime scenes are not entirely free from substrate microbes whose effects on the microbiome-based identification of body fluids are not well understood. In this study, five body fluids (peripheral blood, menstrual blood, nasal secretions, saliva, and semen) were deposited on sterile swabs, bedspreads, and floors under indoor exposure conditions for 7 days. The microbial communities in the samples were characterized using amplicon sequencing targeted V4 region of 16S rRNA gene. The results showed that the microbial communities of fresh samples deposited on sterile swabs clustered together according to the type of body fluid. The microbial composition of the body fluids deposited on the bedspread and floor is significant different from those deposited on sterile swabs. The microbial communities of mock body fluids were a mixture of microbes from pure body fluids and environmental microbes. FEAST analysis showed that the microbes of mock saliva samples were mainly from pure body fluids (51.53 % and 63.04 % on the bedspread and floor, respectively), but not from substrates (25.70 % and 18.92 % on the bedspread and floor, respectively). Contrary results were observed in peripheral blood, mock nasal secretion, and semen samples. All samples were mainly clustered based on the substrate, but not on the type of body fluid in the PCoA visualization. PERMANOVA results showed that the substrate accounted for more of the variance (R2 = 0.211, P < 0.001) than the type of body fluid (R2 = 0.152, P < 0.001). MicroDecon was used to remove contamination by microbes from the substrate of mock body fluid samples. PCoA and PERMANOVA were performed using decontaminated data. The results showed that samples were no longer clustered based on the substrate, and the type of body fluid (R2 = 0.240, P < 0.001) accounted for more of the variance in the microbial communities of samples than the substrate (R2 = 0.108, P < 0.001). Our results suggest that environmental microbiota from substrates may interfere with the microbiome-based identification of forensically relevant body fluids. To some extent, decontamination could decrease the effects of the substrate on the microbial communities of the samples and enhance the ability to distinguish between the types of body fluids. This pilot study will be valuable in promoting the application of microbiome-based stain analysis in forensics.