Background: To perform a study with discriminant power and validity using the Clock drawing Test by instruction (CDTI) in patients with mild Alzheimer's disease (MAD). Materials and Methods: Phase I diagnostic test study. We included a healthy control arm of 58 elderly people and 40 cases with mild Alzheimer's disease. All participants were examined and diagnosed by clinical consensus. The MMSE, CDTI and clinical record were obtained. Results: There were significant differences between the study groups regarding cognitive tests' performance when comparing age and education, but no differences were found when comparing gender. ANCOVA test showed no significant effect exerted by the demographic variables on cognitive performance in any group. The sensitivity (CTO=84[%] vs MMSE=79, 3[%]), specificity (92, 5[%] vs 82, 5[%]) and diagnostic utility of the CDTO were higher than one's resulting form the MMSE (=0, 90, p=0, 000). The combined use of both instruments increased diagnostic capacity. The best cutoff point for the diagnosis of mild dementia was ≤ 6 points in CDTO and ≤ 23 in MMSE. Both instruments correlated statistically. Conclusions: The CTO is a useful test and can discriminate between cognitively healthy subjects and patients with EAL when appliying the "to order" criteria from Cacho's version.