Objective: The Mini-Cog is a simple three part test with a high degree of accuracy (83 percent) that has been found useful in detecting Alzheimer's disease (AD) and related dementias. The test has been shown to identify early dementia in nonnative and non-English speakers as well as in native English speakers. The current study investigated the Mini-Cog's accuracy in differentiating elderly adults with dementia from those without dementia in a Puerto Rican elderly community sample. Method: We selected 50 persons, 60–100 years old, both genders that had a clincial diagnosis of AD certified by a health professional (i.e., neurologist). The control group consisted of 50 elderly volunteers without AD diagnosis, currently living in residential federally funded project homes for the elderly know as “Égidas”. Results: AD (42.9.% females and 57.1% males, MMSE w/ Dx (n = 47; mean 16.32; SD = 5.95), MMSE w/o Dx. (n = 48; mean 12.62; SD = 5.557), ROC MMSE Area: .0774 SE 0.052 p < 0.001, 95% CI Min 0.641 Max 0.846, Mini-Cog and MMSE correlated significant and high (r = .661, p = .000). The Mini-Cog correctly identified (specificity) 96% of subjects-more than either of the other tools. It also had the highest sensitivity at 84% (p < 0.001). The researchers noted that the Mini-Cog's diagnostic value was not influenced by education or language as MMSE is by education and language. Conclusion(s): The results obtained in the present study supported the use of the Mini-Cog in differentiating elderly adults with dementia from those without dementia in a Puerto Rican elderly community sample.