Objective To explore whether repeatedly administered multiple doses of scopolamine to rats is suitable for establishment of an animal model for senile memory deficits or Alzheimer disease research. Methods\ 14 rats were randomly divided into two groups (n=7): normal control group and scopolanmine treated group. In the experimental group, the rats received subcutaneous injection of scopolamine in a dose of 2mg/kg, twice a day for 21 consecutive days. The normal rats received the same volume of saline. Then reference memory testing in Morris water maze(MWM)was followed. Nissl staining was adopted to count the number of pyramidal cells in hippocampal CA1 and CA3 areas under light microscope. Ultrastructural changes of CA1 cells were examined by transmission electron microscopy. Results\ In the experimental group, escape latencies were statistically not significantly different from those of normal control rats on day 1, 2, 4 and 5, except day 3 in place navigation trials. In spatial probe trials, there were no significant differences between two groups. The quantity of pyramidal cells in hippocampal CA1 and CA3 areas was not significantly different between two groups. Ultrastructure of the pyramidal cells including nuclei and cytoplasmic organelles was not much changed. However, the neurons showed a decreased number of synaptic vesicles and reduced post synaptic density (PSD) in scopolamine treated rats in comparison with those of control rats. Conclusions\ Scopolamine given to rats in repeated doses may produce mild impairment of reference memory in Morris water maze. The number and ultrastructure of pyramidal neurons in CA1 and CA3 areas are not seriously changed, but some abnormalities of synapses in hippocampal CA1 cells may occur. Those findings indicate that scopolamine given in repeated doses to rats is not an ideal approach to establish an animal model for Alzheimer disease or senile memory deficits research.