---

We examined influence of the β-amyloid peptide (25-35) neurotoxic fragment (βAP) on Helix lucorum food-aversion learning. Testing with aversively conditioned carrot showed that 2, 5, and 14 days after training the βAP-injected group responded in a significantly larger number of cases and with a significantly smaller latency than the sham-injected control group. The results demonstrate that the amyloid peptide partially impairs the learning process. In an attempt to specify what component of memory is impaired we compared responses in a context in which the snails were aversively trained, and in a neutral context. It was found that the sham-injected learned snails significantly less frequently took the aversively conditioned food in the context in which the snails were shocked, while the βAP-injected snails remembered the aversive context 2 days after associative training, but were not able to distinguish two contexts 5, and 14 days after training. In a separate series of experiments a specific context was associated with electric shock, and changes in general responsiveness were tested in two contexts several days later. It was found that the βAP-injected snails significantly increased withdrawal responses in all tested contexts, while the sham-injected control animals selectively increased responsiveness only in the context in which they were reinforced with electric shocks. These results demonstrate that the β-amyloid peptide (25-35) interferes with the learning process, and may play a significant role in behavioral plasticity and memory by selectively impairing only one