Accurate 3D flow characterization in a microchannel is becoming increasingly important for the design and development of microfluidic chips. In recent years, a light field camera that can simultaneously record the direction and position information of rays in a single photographic exposure has been developed and employed in the field of computer graphics. In this paper, a microparticle image velocimetry based on light field imaging (light field $\mu $ PIV) is proposed to reconstruct the 3D velocity field of a microscale flow. Both simulations and experiments are performed to verify the proposed method. The light field image of tracer particles and the point spread function (PSF) of a light field microscopic imaging system are numerically calculated based on the Abbe imaging principle. The 3D positions of the tracer particles in a flow field are then reconstructed by the Lucy–Richardson 3D deconvolution algorithm. Furthermore, a light field $\mu $ PIV system based on an assembled cage light field camera with a microscope is developed, and calibrations are performed to obtain the geometric parameters of the $\mu $ PIV system accurately. The simulation and experimental results demonstrate the feasibility of the proposed light field $\mu $ PIV. Compared with the synthetic refocusing reconstruction method, the Lucy–Richardson 3D deconvolution algorithm greatly improves the lateral and the axial resolutions of the flow field.