Although the immunotherapy for Alzheimer’s disease (AD) has achieved some exciting results, current immunization strategies with amyloid β-protein (Aβ) and its antibodies are causing serious side effects in AD patients, which is partly due to the damage of physiological amyloid precursor protein. Our previous studies indicated that the sequence 31-35 in Aβ is a shorter active center responsible for the neurotoxicity of Aβ, suggesting a more specific target for AD treatment. The present study, on the basis of making a specific antibody against the sequence 31-35 of Aβ, observed the effects of the anti-Aβ31-35 antibody on the Aβ1-42-induced impairments of spatial learning and memory in rats and investigated its electrophysiological and cellular mechanisms using Morris water maze (MWM), primary neuronal culture and hippocampal field potential recording. Under the anesthesia with 1% carbrital (50 mg/kg, i.p.), a stainless steel cannula was implanted into the right lateral ventricle of Wistar rats (200-300g) for intracerebroventricular (i.c.v.) injection of drug/vehicle. MWM tests consisted of acquisition phase and probe trial. The hippocampal long term potentiation (LTP) was examined by recording field excitatory postsynaptic potentials (fEPSPs) in the CA1 region of rats under anesthesia with ethyl carbamate (1.5 g/kg, i. p.). In primary culture experiments, cell viability, LDH release and cell survival rate were checked. All data were presented as means ± standard error of the mean. The statistical significance was defined as p<0.05. The results showed that i.c.v. injection of anti-Aβ31-35 antibody alone did not change the spatial memory (p>0.05, n=8), the basal synaptic transmission and hippocampal LTP (p>0.05, n=7) of rats. However, pretreatment with the anti-Aβ31-35 antibody dose-dependently (0.05-5 nmol) decreased the escape latencies in underwater platform test (p<0.05 in 0.5 and 5 nmol group, n=8) and increased the swim time and distance elapsed in the target quadrant in probe trial (p<0.05, n=8); 5 nmol Anti-Aβ31-35 antibody almost completely reversed the Aβ1-42-induced suppression of in vivo hippocampal LTP (p<0.01, n=6); Aβ1-42-induced cytotoxicity in cultured cortical neurons was also inhibited by the antibody (p<0.05 or p<0.01), with a nearly normal cell viability and a reduced LDH release. These results suggest that 31-35 sequence of Aβ may be a new therapeutic target for AD, and the anti-Aβ31-35 antibody will be probably useful in AD immunotherapy as a novel anti-Aβ antibody candidate.