Particle image velocimetry (PIV), a technique commonly used at macroscopic length scales to measure velocity fields of particle-seeded flows, is adapted to measure velocity fields in microfluidic MEMS devices, where micron-scale spatial resolution is critical. Adapting PIV to the microscopic level presents a number of challenges, including visualizing tracer particles that are smaller than the wavelength of light and minimizing errors due to the Brownian motion of the tracer particles. High numerical aperture video microscopy is used to record the faint signals from fluorescent 300 nm particles. Innovative ensemble averaging and adaptive spatial shifting algorithms are used to extract maximal information from the recorded images. The PIV technique is used to measure a low Reynolds number Hele-Shaw flow around an 8 μm human red blood cell. The velocity vector field presented has a maximal spatial resolution of 3.2 × 3.2 × 1.5 μm.