Dedicated small animal positron emission tomography (PET) systems are increasingly prevalent in industry (e.g. for preclinical drug development) and biological research. Such systems permit researchers to perform animal studies of a longitudinal design characterised by repeated measurements in single animals. With the advent of commercial systems, scanners have become readily available and increasingly popular. As a consequence, technical specifications are becoming more diverse, making scanner systems less broadly applicable. The investigator has, therefore, to make a decision regarding which type of scanner is most suitable for the intended experiments. This decision should be based on gantry characteristics and the physical performance. The first few steps have been taken towards standardisation of the assessment of performance characteristics of dedicated animal PET systems, though such assessment is not yet routinely implemented. In this review, we describe current methods of evaluation of physical performance parameters of small animal PET scanners. Effects of methodologically different approaches on the results are assessed. It is underscored that particular attention has to be paid to spatial resolution, sensitivity, scatter fraction and count rate performance. Differences in performance measurement methods are described with regard to commercially available systems, namely the Concorde MicroPET systems P4 and R4 and the quad-HIDAC. Lastly, consequences of differences in scanner performance parameters are rated with respect to applications of small animal PET.