This study aimed at developing a rat model of obsessive compulsive disorder based on the hypothesis that a deficient response feedback mechanism underlies obsessions and compulsions. Rats were trained to lever press for food, whose delivery was signaled by the presentation of a compound stimulus (light+tone). Subsequently, the classical contingency between the stimulus and food was extinguished (signal attenuation). Experiment 1 showed that this manipulation resulted in increased lever pressing during a subsequent extinction test, which was highly correlated with an increase in the number of trials on which the rat did not attempt to collect the food reward. This behavioral pattern was not evident in an extinction test not preceded by signal attenuation (Experiment 2), suggesting that the latter is a crucial factor in the development of this behavioral pattern. Excessive lever pressing was attenuated by the selective serotonin re-uptake inhibitor, fluoxetine (10 mg/kg; Experiment 3), but not by the anxiolytic drug, diazepam (2 mg/kg; Experiment 4). Based on these results we propose that post-training signal attenuation may provide a rat model of obsessive compulsive disorder.