Surface enhanced Raman scattering (SERS) and surface enhanced resonance Raman scattering (SERRS) from a silver colloid suspension are compared using a dye designed to bond strongly to a silver surface. Titration of the dye into a silver colloid suspension caused aggregation in a controlled manner without an aggregating agent being added. Concentrations of dye equivalent to between 5 x 10(-9) and 10(-3) M in the final dilution before adsorption to the silver were used. The results suggest that nionolayer coverage of the surface occurs at approximately 10(-6) M. Above this concentration, the suspensions are less stable, and the relationship between intensity and concentration is complex. Below this concentration, three main regions can be identified by electronic absorption spectroscopy. At dye concentrations up to 7.5 x 10(-8) M, there is little evidence of aggregation, although there are changes in the spectra ascribed here. to surface changes caused by dye adsorption. Between 7.5 x 10(-8) M and 2.5 x 10(-7) M, a well-defined small aggregate appears to occur and above 2.5 x 10(-7) M larger less well-defined aggregates form. SERRS gave linear concentration dependence below 7.5 x 10(-8) M suggesting scattering from single particles. A changeover region occurs close to where the first evidence of aggregation is detected by electronic spectroscopy, and at higher concentrations up to about monolayer coverage a second linear region was obtained. SERS below the concentration at which aggregation is detected by electronic spectroscopy was weak and difficult to obtain. At higher concentrations, the SERS gradients are steeper and the maximum enhancement observed is within a factor of 4 of that obtained in SERRS. The study shows that there is a different mechanism in SERRS compared to SERS with single particle enhancement being much greater in SERRS.