In 3 experiments discontinuous shock (shock having brief on-periods and longer off-periods) was found to facilitate, often markedly, discriminated bar-press avoidance learning in the rat. The type of CS employed (light, white noise, or a combination of the 2) was also found to be a relevant variable, with a strong suggestion that this variable interacts withthe shock variable. The nature of the interaction is that, although continuous shock results in poorer avoidance performance with all 3 CSs, the combination of light CS and continuous shock is detrimental to the development of avoidance performance.