The search for the “Holy Grail” in clinical diagnostic microbiology—a reliable, accurate, low-cost, real-time, easy-to-use method—has brought up several methods with the potential to meet these criteria. One is Raman spectroscopy, an optical, nondestructive method based on the inelastic scattering of monochromatic light. The current study focuses on the possible use of Raman spectroscopy for identifying microbes causing severe, often life-threatening bloodstream infections. We included 305 microbial strains of 28 species acting as causative agents of bloodstream infections. Raman spectroscopy identified the strains from grown colonies, with 2.8% and 7% incorrectly identified strains using the support vector machine algorithm based on centered and uncentred principal-component analyses, respectively. We combined Raman spectroscopy with optical tweezers to speed up the process and captured and analyzed microbes directly from spiked human serum. The pilot study suggests that it is possible to capture individual microbial cells from human serum and characterize them by Raman spectroscopy with notable differences among different species.