We read with great interest the review by Nitrini et al. on the prevalence of dementia in Latin America recently published in International Psychogeriatrics (Nitrini et al., 2009). Accurate up-to-date figures are essential for policy-making and planning, therefore the review is very welcome. With unfortunate timing, the 10/66 Dementia Research Group's population-based surveys on the prevalence of dementia were published in the Lancet (Llibre Rodriguez et al., 2008a; 2008b) shortly after this review was submitted to International Psychogeriatrics. The 10/66 surveys included seven sites in five Latin American countries: Peru, Cuba, Dominican Republic, Mexico and Venezuela. The studies were all one-phase catchment area surveys, with samples of 2944 in Cuba and between 1904 and 2011 in other countries, giving a total sample size of 10,794. We present in Table 1 the prevalence of dementia according to our cross-culturally validated 10/66 diagnosis and according to DSM-IV criteria, in each of the Latin American sites, using the same age group stratification as per Nitrini's review. We also present the pooled estimates for each age group. The 10/66 estimates are in general more homogenous than those presented in the review, but similar to the overall pooled estimate. DSM-IV prevalence is lower. We have attributed this discrepancy to an under-reporting of cognitive decline and social/occupational impairment by relatives, particularly in rural and least developed regions (Llibre Rodriguez et al., 2008b). We have shown that, at least for Cuba, the 10/66 Dementia Diagnosis agreed better than the DSM-IV with a clinician gold standard diagnosis, as a high proportion of Clinical Dementia Rating mild and moderate cases were missed by DSM-IV (Prince et al., 2008).