To investigate alternation behavior in the cockroach as an invertebrate, three T-maze experiments were conducted assessing the effects of (a) intratrial (exposure) and intertrial (exposure to test) interval, (b)brightness similarity of the alternatives and response-direction factors, and (c) an initial free-choice vs. a varying number of forced-choice exposures. Alternation was enhanced with a long exposure on Trial l , a short interval between Triall and Trial 2 (test), greater dissimilarity (or discriminability) of the arm brightnesses and more frequent forced-choice exposures to an arm prior to test. These. results highlight the role of intramaze brightness cues as controlling alternation in the cockroach and accord well with a response-to-change interpretation of the phenomenon. In addition, they suggest that the alternation paradigm can be used effectively to assay short-term memory storage in neurologically simple organisms. Schultz, 1964). Apart from investigations employing rat and human subjects, however, the research with other organisms has served mainly to document the occurrence of alternation, for there has been little study of variables controlling the phenomenon.The purpose of the present study was to extend investigation of the phenomenon in lower organisms by assessing factors affecting alternation in the cockroach as an invertebrate. Such an investigation is particularly important when it is considered that recent research with the rat (e.g., Bronstein, Dworkin, Bilder, & Wolkoff, 1974;Douglas, 1966; Eisenberger, Myers, Sanders, & Shanab, 1970) has contested the basis on which alternation occurs. Using the rat literature as a guide, the present study investigated six variables treated two at a time in three independent experiments which were ordered so as to facilitate parameter selection. In succession, the three experiments assessed the effects of: (1) intratrial (exposure) and intertrial (exposure to test) interval, (2) brightness similarity of the choice alternatives and response-direction factors, and (3) an initial free-vs. a varying number of forced-choice exposures.