The study reported here demonstrates the feasibility of presenting cognitive tests from the Cambridge Neuropsychological Test Automated Battery (CANTAB) to either marmosets or rhesus monkeys in their home cages. This location of testing offers opportunities for the measurement of additional indices, for example spontaneous behaviour (Prowse et al. 1995) and electrophysiology (Pearce et al. 1998) as well as facilitating repeated test presentation. Results from 12 marmosets and 4 rhesus monkeys which have completed several sequences of an eight-stage discrimination task involving simple discriminations, compound discriminations and reversals are reported. The paradigm developed has application in long-term studies. Tests from CANTAB have been used extensively in normal humans (Robbins et al. 1994) as well as a range of patient groups (Owen et al. 1992, Elliott et al. 1995) and to assess drug effects (Coull et al. 1996). Additionally some of these tests have been presented to marmosets (Roberts et al. 1988) to examine neuropsychological functioning. This comparative approach facilitates meaningful cross species comparison, particularly in the study of the effects of pharmacological intervention.