Separate groups underwent fear conditioning (tone‐shock trials) with CS‐UCS intervals favouring reinforcement (CR‐UCS offset; Group 1), CS‐UCR contiguity (Group 2), and contiguity and reinforcement (interval shifted during training; Group 3) while a control group (Group 4) had unpaired presentations of CS and UCS. Fear conditioning, extinction, and spontaneous recovery were measured through the probestimulus technique, with light as probe stimulus and pre‐instructed hand lifting as probe response. Predictions derived from Jones' (1962) hypothesis were confirmed; Groups 2 and 3 were superior to Group 1 at first, while Group 3 was superior to Group 2 later in training. All experimental groups showed extinction, and Groups 1 and 3 showed spontaneous recovery in later retraining. The data were taken to confirm the main hypothesis that the fear response itself has a latency of the order of 2000