Unpolarized light from the central star that is reflected by exoplanets, protoplanetary disks, and debris disks becomes partially polarized by the reflection process. Imaging polarimetry is therefore the ideal way to discriminate between the polarized light from circumstellar environments and the unpolarized light from the nearby central star. A sensitivity of 10 $^-5$ (fraction of polarized intensity to the total intensity) must be achieved to detect exoplanets; 10 $^-4$ is sufficient for disks. Based on extensive experience in precision polarimetry of the Sun, the newly formed experimental astrophysics group at Utrecht University, The Netherlands, will design, build, and use a high-precision imaging polarimeter for use at the 4.2-meter William Herschel Telescope. Since systematic errors typically limit conventional imaging polarimeters to about 10 $^-3$, laboratory setups and theoretical models will be used to understand and then minimize and/or calibrate systematic errors. Published catalogues of exoplanets and stars that harbor disks will guide extensive observations with this new polarimeter. The effort will focus on retrieving fundamental properties of circumstellar environments that cannot be obtained with other observational approaches.