Scrublet is a computational framework designed to address the issue of cell multiplets in single-cell RNA-sequencing (scRNA-seq) data. Multiplets, where multiple cells are labeled with the same barcode, can confound downstream analyses by introducing artifacts and misleading interpretations. Scrublet avoids the need for expert knowledge or cell clustering by simulating multiplets from the data and building a nearest neighbor classifier. The method is validated on simulated and real datasets, demonstrating its ability to accurately identify and classify neotypic (newly formed) and embedded (existing) multiplets. Scrublet provides a posterior likelihood of a cell being a doublet, allowing for the removal of problematic multiplets and improving the quality of downstream analyses. The implementation is available online, making it a valuable tool for researchers working with scRNA-seq data.Scrublet is a computational framework designed to address the issue of cell multiplets in single-cell RNA-sequencing (scRNA-seq) data. Multiplets, where multiple cells are labeled with the same barcode, can confound downstream analyses by introducing artifacts and misleading interpretations. Scrublet avoids the need for expert knowledge or cell clustering by simulating multiplets from the data and building a nearest neighbor classifier. The method is validated on simulated and real datasets, demonstrating its ability to accurately identify and classify neotypic (newly formed) and embedded (existing) multiplets. Scrublet provides a posterior likelihood of a cell being a doublet, allowing for the removal of problematic multiplets and improving the quality of downstream analyses. The implementation is available online, making it a valuable tool for researchers working with scRNA-seq data.