This study investigates a method to predict medical outcome of cholinesterase inhibitors in patients with Alzheimer's disease (AD) and vascular dementia (VaD). Van Gool predicts that patients with cholinergic deficit symptoms will benefit from treatment whereas patients without will experience side effects because of overstimulation of the cholinergic system. We predicted that AD and VaD patients with a longer RT experience fewer side effects than patients with a faster response and that VaD patients have a longer RT than AD patients. A number of 71 patients with AD or VaD diagnosis were included. A sustained attention task was administered, as well as the MMSE and a questionnaire about side effects. Results indicated that VaD patients with a longer RT reported fewer side effects. Furthermore, patients with VaD had a longer RT than patients with AD. MMSE was negatively correlated with RT in the VaD group. Thus, the performance on the attention task seems associated with suffering from side effects and thus tends to predict medical outcome in VaD, but not in AD. Perhaps this attention task was not sensitive enough to measure cholinergic deficit symptoms in AD patients. Furthermore, different doses of medication might confound the effect for the AD group.