62 rats, operationally divided into those who were seizure-prone and nonseizure-prone, gave similar activity measures over the shorter delays of test but seizure Ss in the longer delay groups moved less than nonseizure Ss before and after bell-can treatment. This finding was interpreted as evidence for an interaction between seizure-prone condition with longer periods of fasting. When Ss were tested in a 4-manipulandum test situation, reinforcement schedules (CRF, FR-4, VR-4, and controls) were differentially effective. Preferences for specific devices developed. Response variability did not differentiate seizure and nonseizure Ss but was significantly influenced by schedules of reinforcement.