Nano-particle image velocimetry (nPIV) uses evanescent-wave illumination to measure two velocity components, U and V, tangent to the wall in a region with thickness of order of hundred nano meters. In this region the illumination intensity decays exponentially with distance normal to the wall, z, and hence tracers closer to the wall have brighter and bigger images than those that are further away, i.e. at larger z. Moreover fluid velocity varies in this region with z and hence tracers at different distance from the wall move at different speeds. Furthermore, Brownian displacement of particle tracers in this region is comparable to the displacement due to the fluid convection. The variation in the displacement of particle images in this region, with different brightness and velocities, can bias the near-wall velocities obtained using standard correlation based PIV method. Artificial nPIV images of nano particle in a flow field with linear out of plane velocity profile were used in this work to investigate the impact of these issues upon the accuracy of nPIV data. Uniform and Gaussian random distribution noise were added to the images to simulate electronic noise and shot noise, respectively. The artificial images were obtained and processed for various experimental parameters to incorporate different illumination profile and shear rates. The results demonstrate that non-uniform illumination affects the bias in the estimated tracer velocity for the shear flow. Non-uniform intensity also affects the bias due to Brownian diffusion; however, correction for Brownian diffusion can reduce this bias error.