Objectives: To establish a statistical relationship between poverty ad perinatal mortality at the individual level (death event) or contextual level (surrounding data). Patients and Method: Deaths of newborns in the 27 days of life were examined. The period between Jan 1st and Dec 31 of each year between 1996 and 2005 was considered. Individual data was verified by: gender, cause of death (ICD-10) and perinatal cause of death (P00-P96), ethnicity, age, district, community of procedence. Results: Poisson regressions showed statistical significance a < 0.05 for the relationship between poverty and perinatal mortality. This included smoothing for more stable estimation. All lineal general models showed statistical significance. Higher poverty level correlates with higher perinatal death. Chi Squared Scaled Pearson decreases a standard error in the interception estimator and the dependent variable (poverty). Conclusions: There is increased evidence that poverty and material risk have a clear effect upon population mortality, especially on vulnerable age groups. Ethnic variables show that poverty levels are significantly higher among Mapuche Population.