Conventional charged-particle imaging techniques--such as autoradiography--provide only two-dimensional (2D) images of thin tissue slices. To get volumetric information, images of multiple thin slices are stacked. This process is time consuming and prone to distortions, as registration of 2D images is required. We propose a direct three-dimensional (3D) autoradiography technique, which we call charged-particle emission tomography (CPET). This 3D imaging technique enables imaging of thick sections, thus increasing laboratory throughput and eliminating distortions due to registration. In CPET, molecules or cells of interest are labeled so that they emit charged particles without significant alteration of their biological function. Therefore, by imaging the source of the charged particles, one can gain information about the distribution of the molecules or cells of interest. Two special case of CPET include beta emission tomography (BET) and alpha emission tomography (𝛼ET), where the charged particles employed are fast electrons and alpha particles, respectively. A crucial component of CPET is the charged-particle detector. Conventional charged-particle detectors are sensitive only to the 2-D positions of the detected particles. We propose a new detector concept, which we call particle-processing detector (PPD). A PPD measures attributes of each detected particle, including location, direction of propagation, and/or the energy deposited in the detector. Reconstruction algorithms for CPET are developed, and reconstruction results from simulated data are presented for both BET and 𝛼ET. The results show that, in addition to position, direction and energy provide valuable information for 3D reconstruction of CPET. Several designs of particle-processing detectors are described. Experimental results for one detector are discussed. With appropriate detector design and careful data analysis, it is possible to measure direction and energy, as well as position of each detected particle. The null functions of CPET with PPDs that measure different combinations of attributes are calculated through singular-value decomposition. In general, the more particle attributes are measured from each detection event, the smaller the null space of CPET is. In other words, the higher dimension the data space is, the more information about an object can be recovered from CPET.