Background The results showed that mirror writing (MW) was correlated with the development of written language, so that MW examination may be one of methods to examine the intelligence of elderly people. Objective To study the MW in elderly patients with Alzheimer disease (AD) and vascular dementia (VaD) and take appropriate scale for their evaluation. Design Taking the written portion of the Chinese Aphasia Examination Scale (1994) for assessment. Setting Department of Neurology, Neuropsychological Laboratory, Beijing Hospital. Participants From March 1998 to January 2001, 33 patients with AD, 30 patients with VaD admitted into Department of Neurology, Beijing Hospital was enrolled into study. Criteria according to the Diagnostic and Statistical Manual of Mental Disorder, 4 th edition (DSM-IV), published by the American Psychiatric Association was used to diagnose AD, while criteria according to the National Institute of Neurological Disorders and Stroke-Association Internationale pour la Recherche et l' Enseignement en Neurosciences (NINDS-AIREN) and Alzheimer Disease Diagnostic and Treatment Center (ADDCT) were used for diagnosis of VaD. AD group contained 19 males and 14 females aged 60–83 years. Twenty-eight males and 2 females, aged 60–87 years made up the VaD group. The 63 healthy elderly subjects matched on age and education as controls were enrolled into study. The matched controls were categorized AD control ( n =33) and VaD control ( n =30). All patients and controls were understanding and agree with all items of assessment. Methods MW examination, Mini Mental State Examination (MMSE), Hachinski Ischemic Scale, the Global Deterioration Scale (GDS) were examined in all subjects. ▪ Use the written potion of the Chinese Aphasia examination Scale (1994), patient using MW for 91%–100% of dictation had complete MW, those using MW for 51%–90% of dictation had severe MW, those using MW for 11%–50% had moderate MW, those using MW for 1%–10% had mild MW. ▪ According to the MMSE, the patients were considered to have dementia if they were illiterate and had an MMSE score = 17 score or educated time = 6 years and MMSE = 20 score or educated time > 6 years and MMSE = 24 score. ▪ Using the Hachinski Ischemic Scale to differentiate the AD and VaD, it included 6 items and total 9 scores. > 7 score was VaD, < 4 score was AD and 4-7 score was blended dementia. ▪ Using the GDS to assess cognitive function: Standard criteria were divided in 7 degrees: 1 degree: no impairment of cognition and 7 degree: very severe impairment of cognition. Main outcome measures The data of MW examination, evaluation of MMSE, Hachinski Ischemic Scale and the GDS of all assessed subjects. Results All 63 cases of AD and VaD and 63 healthy controls were entered for analysis. ▪ Results of MW examination: A total of 17 patients with AD were characterized as using MW, 3 with moderate MW and 14 patients with mild MW. In the corresponding control group, only 2 subjects were characterized as being mild MW. The VaD group has 23 patients with MW, 2 with moderate and 21 patients with mild MW. ▪ MMSE score: MMSE score of AD group was much lower than that of individuals in control group [(20.15±3.40), (29.73±0.40) score, P < 0.01], MMSE score of VaD group was much lower than that of individuals in control group [(19.33±2.75), (29.12±0.63) score, P < 0.01]. ▪ Hachinski Ischemic Scale and GDS score: Rating according to the Hachinski Ischemic Scale was higher in VaD patients compared to AD patients [(9.61±1.69), (1.09±0.60) score, P < 0.01]. The GDS score did not significantly differ between the AD group and the VaD group. Conclusion ▪ MW examination could be used as an indicator of intelligence in healthy elderly people and also could be used as one of methods to assess the intelligence in AD and VaD patients. ▪ Grade of severity of MW may indirectly reflect the degree of dementia.