Amnesia for a single trial discriminated passive avoidance learning task in day-old chicks, resulting from the antibiotic cycloheximide, is shown to begin after 30 min following learning when cycloheximide is given immediately after or before learning. However, amnesia does not begin until after 50 min following learning when cycloheximide is given 5 min or later after learning. The results are interpreted in the context of a 3 stage model of memory formation. It is suggested that the second or intermediate stage involves two phases: phase A which lasts up to 30 min following learning, is energy dependent, and is susceptible to inhibition by 2,4-dinitrophenol; phase B which occurs after phase A, lasts up to 50 min following learning, is energy independent, and is not susceptible to inhibition by 2,4-dinitrophenol. It is concluded that cycloheximide given immediately after or before learning inhibits expression of phase B of intermediate memory as well as formation of long-term memory, while cycloheximide given 5 min or later after learning only inhibits formation of long-term memory. Another antibiotic, anisomycin, is shown to have no effect on intermediate memory, independent of time of administration.