A double blind trial on the effects of 2 mg of benzhexol on memory was performed on thirteen elderly subjects without cognitive impairment. The tests given 90 minutes after the drug or placebo involved learning a list of ten words, a paired-associate learning task, learning a short story and a test of digit span. Subjects were asked to recall the word list one minute after an interfering task, and 6 items from the story directly. Digit span involved immediate recall and the paired learning was measured by the number of trials necessary to learn. All tests were significantly impaired by the benzhexol except for digit span. This suggests that muscarinic blocking drugs should be avoided in the elderly, as they mimic the memory deficits found in senile dementia of Alzheimer type.