CASToR is an open-source multi-platform project for 4D emission (PET and SPECT) and transmission (CT) tomographic reconstruction. This platform is a scalable software providing both basic image reconstruction features for “standard” users and advanced tools for specialists in the reconstruction field, to develop, incorporate and assess their own methods in image reconstruction (such as specific projectors, optimization algorithms, dynamic data modeling, etc) through the implementation of new classes. A generic and flexible input data file format has been designed in order to integrate all the information needed for the reconstruction, for any modality and for any data format (list-mode or histogram). Being strictly a reconstruction software, and due to its generic design, CASToR will not estimate correction factors specific to each data channel (namely normalization, attenuation, scattered and random counts, etc.), so all these corrections should be pre-computed and embedded in the data file. The aim is to design a generic core algorithm flexible enough to cope with the particularities of each imaging modality through a user-friendly class system allowing for the integration of new methods. CASToR is highly portable on different platforms and it does not depend on any library for a standard use (i.e. on a single multi-core computer). The compilation of the source code is fast, straightforward and does not require being a computer geek. Binaries compiled for Unix and Windows 64-bits systems and Windows 32-bits systems are also provided.