Multi-wavelength transmission spectrum can reflect the unique information on cell size, shape, internal structure and chemical composition of a sample, so multi-wavelength transmission spectroscopy is a powerful technique for real-time and in-site detection and identification of cells. And the development of multi-wavelength transmission spectroscopy for the rapid and effective detection of bacterial microbes in water will be of great significance to the control of microbial contamination in water and the protection of the safety and health of water quality. In this paper, in order to develop the method of rapidly and accurately detecting the pathogenic bacterial microbes in water based on multi-wavelength transmission spectroscopy, the multi-wavelength transmission spectra in the range of 200～900 nm of various pathogenic bacterial microbes such as Klebsiella pneumoniae, Salmonella typhimurium, Staphylococcus aureus and Escherichia coli were obtained adopting UV-Vis spectrophotometer for the first time, respectively. And the spectral features of different bacteria and the same bacteria with different concentrations were compared and analyzed. The results demonstrate that for different bacteria, when the concentration changes, the spectral shapes in the range of 400～900 nm are consistent, and the optical density value at 400, 450, 500 and 550 nm respectively has a very good linear relationship with bacterial concentration. But in the range of 200～400 nm, the spectral shapes change with the change of bacterial concentration, and the optical density value at 200, 258, 300 and 350 nm respectively has a very good quadratic polynomial relationship with bacterial concentration. According to Mie scattering theory, Levenberg-Marquardt nonlinear least square method was adopted to calculate the scattering spectra and absorption spectra based on the measured transmission spectra of different bacteria. And the spectral features of normalized scattering spectra and absorption spectra of different bacteria were compared and analyzed. The results demonstrate that for the four types of bacteria, all the characteristic spectral peaks of scattering spectra are at 245 nm. But for different bacteria, the optical density values at characteristic peak are obviously different, which arises from the differences in cell size and shape of different bacteria. Furthermore, all the characteristic spectral peaks of absorption spectra of the four types of bacteria are at 260 nm. But for different bacteria, the spectral absorption band between 240～400 nm and spectral intensity at 260 nm are obvious different, which are attributed to the difference in content of chemical composition such as nucleic acid and protein in different bacteria. This study indicates that for the different bacteria and the same bacteria with different concentrations, the spectral features of multi-wavelength transmission spectra, calculated scattering spectra and absorption spectra are obviously different. And various parameters of bacteria can be obtained by the interpretation of multi-wavelength transmission spectra. So multi-wavelength transmission spectroscopy can be used to rapidly and effectively detect pathogenic bacterial microbes in water. This study provides an important basis for the development of rapid and on-line monitoring instrument of bacteria in water.