In this work, surface-enhanced Raman spectroscopy (SERS) method based on uncoated spherical gold nanoparticles as SERS substrate was applied for the detection and discrimination of three different bacterial strains: Gram-positive Bacillus subtilis and Lactobacillus rhamnosus GG and Gram-negative Escherichia coli. The exploitation of bacteria SERS spectra by multivariate data analysis (principal component analysis and partial least squares-discriminant analysis) proved the robustness of the developed method and its suitability to discriminate between the three different strains despite their relatively similar Raman signals. The SERS performance of the developed method was compared with that of the classical silver nanoparticles indicating a better bacteria signal enhancement, up to 15 times higher in case of E. coli, likely due to stronger interactions of gold nanoparticles with components such as purine molecules that show dominant bands in the bacterial SERS spectra. K E Y W O R D S bacteria detection, bacteria discrimination, gold nanoparticles, SERS