Behavioral paradigms applicable for use in both human and nonhuman subjects for investigating aspects of working/short-term memory are presented with a view towards exploring their strengths, weaknesses, and utility in a variety of experimental situations. Such procedures can be useful in teasing out specific aspects of mnemonic processes including discrimination, encoding, and retention. Delayed matching-to-position, delayed matching-to-sample (DMTS), and titrating matching-to-sample procedures are highlighted. Additionally, the application of DMTS tasks in preclinical and clinical settings is presented: drug effects on memory processes can be explored preclinically in animal models; normative data have been developed in human populations where they have been used in adults to explore the relationships between mnemonic processes and specific clinical entities such as Parkinsonism, senile dementia of the Alzheimer's type, schizophrenia, and depression. Studies in children indicate that encoding and retention processes improve rapidly in the early years, plateauing prior to puberty. Noninvasive imaging techniques such as positron emission tomography (PET) indicate that activity in specific brain areas is associated with DMTS task performance and may serve to confirm roles for such structures in mnemonic processes.