Raman spectroscopy and its counterpart surface-enhanced Raman scattering (SERS) have proven to be effective methods for detecting miniscule changes in the phenotypes of E. coli and other single-celled organisms to aid in the detection of new strains for industrial use and discovery of new antibiotics. The purpose of this study is to develop a method to quickly and accurately detect contaminants in water samples through phenotype changes in E. coli measured through SERS. Contaminated Luria-Bertani (LB) media was inoculated with LB with an OD600 of 1, grown for two hours, and then dried on a flat piece of aluminum foil. These samples were then Raman scanned and processed to determine contaminant-induced changes to the phenotypes of the E. coli. Three types of tests were run to show the effectiveness of this method: single-component, multicomponent, and impure water sources. In single-component tests, it was found that differences due to NaCl contamination could be detected to 5.0E-9 weight percent (wt %), ethanol (EtOH) to 5.0E-7 volumetric percent (% v/v), citric acid (CA) to 2.8E-4 wt %, acetic acid (AA) to 2.6E-4 wt %, kanamycin to 2.5E-11 wt %, ampicillin to 2.5E-10 wt %, CoCl2 to trace amounts, and silver nanoparticles (AgNP) to 5.2E-7 wt %. Many of these are below the detection limits of analytical instrumentation, but their effects on E. coli phenotypes were detectable by Raman spectroscopy. Multicomponent tests showed that in a mixture, the most toxic or most concentrated contaminants have the most effect on cell phenotype. However, it was shown that similar concentrations of similar contaminants may be difficult to discern with current methods. This behavior was also seen in the impure water samples, showing that tap water behaves the closest to a DI control, followed by running water, and finally stagnant bodies. This new method of monitoring E. coli phenotypes with Raman spectroscopy as a biosensor shows promise for the fast, portable, and accurate determination of environmental contaminants with a broad-spectrum and very low detection limits.