INTRODUCTION The health impacts of environmental risk factors are complex to evaluate. Confounding and synergies can occurs and the relationship is not often linear. In the context of climate change, studies on the health impacts of temperature should observe this dynamic effects and interactions, including controlling for air pollution exposure. Studies based on short time series can address this impacts but its generalisation the climate change scenario is made with reservations. METHODOLOGY This study presents a time series analysis (Poisson regression) based on 30 years (1975 - 2005) of mortality and environmental daily data for Los Angeles-CA. The long term trends and seasonality were controlled. RESULTS The cardiovascular mortality presented a expressive negative correlation with temperature, both minimum and maximum in the wet and dry season. The respiratory mortality has a negative correlation only for the wet season. Precipitation didn't present representative association with the health outcomes. The effects of air pollution were detected and controlled. CONCLUSIONS The climate change scale is greater than the most common time series analysis available. This study presented one of the pioneer initiatives to evaluate the effect on mortality of meteorological variable, such temperature, on the health based on a 30 years time series. The results provide a better understanding of the complex association dynamics, including the effect of air pollution.