Using Au nanoparticles, it is possible under certain experimental conditions to considerably enhance the sensitivity of a conventional surface plasmon resonance (SPR) device. In this report, we examine the mechanism of this enhancement and discuss the experimental factors that are crucial to the performance of such a nanoparticle based SPR device. Among these factors, the surface plasmon-supporting metal substrate plays a major role. We demonstrate this by comparing experimental SPR data for Au and Ag substrates. In both cases, ∼25-30 nm diameter Au particles are attached to the substrate metal via a sandwiched monolayer of 1,6-hexanedithiol. The width and the efficiency of SPR for both substrates are affected by the Au particles. However, the particle-induced shift in the SPR angle is observed only for the Au substrate. This observation is explained in terms of competitive effects of propagating surface plasmons in the substrate metal and localized surface plasmons in the Au nanoparticles.In recent years, the SPR imaging technique has been successfully employed for immunosensing and for thickness measurements of ultrathin (2-20 Å) films. [1][2][3] In a typical SPR device, the sample layer is adsorbed onto a metal film, and the total attenuated reflection (ATR) efficiency, R, of the metal is measured as a function of the incidence angle, θ, of a probe light beam. The behavior of R is determined by the propagating surface plasmon polariton (SPP) modes at the metal/sample interface. In the absence of the sample, the R-θ plot (SPR plot) exhibits a dip at the angle, θ p , where the condition for SPR (resonance in SPP oscillations) is satisfied. In the presence of the sample layer(s), θ p shifts to a new value, and this shift is analyzed to determine the thickness (amount) and/or the dielectric function (chemical properties) of the sample. 4,5 Usually, the width, position and height of the SPR plot, as well as the fractional reflectivity change, (∆R/R) p , at the SPR angle are also modified in the presence of the adsorbed sample. These latter changes can provide additional information about the sample. However, often the sensitivity of the SPR device is limited by small shifts in θ p and (∆R/R) p . An effort to overcome the limitations of conventional SPR imaging have been recently reported by Lyon et al. 6,7 The discussion of our present communication focuses on the SPR device proposed by these authors. They use a gold nanoparticle-modified device, where an organic self-assembled monolayer (SAM) is sandwiched between a gold film and a layer of colloidal gold particles (in the 10-60 nm diameter range). The experimental sample is adsorbed on top on to the Au nanoparticles. The Au particles enhance the magnitude of ∆θ p , and at the same time, introduce an additional change in (∆R/R) p . These particles have their own