This research aims to analyze and estimate future scenarios of maximum air temperature in the capitals of northeastern Brazil, in order to highlight the importance of climate change today and in the future. For this, rainfall, wind speed, relative humidity and maximum air temperature data were used by the database meteorological activities of the National Institute of Meteorology, of the nine capitals of the northeastern region of Brazil from 1980 to 2019, and the dynamic regression technique that combines the dynamics of time series and the effect of explanatory variables.The main results showed that the dynamic regression model satisfactorily adjusted the association between meteorological variables.Trend (without lag) and seasonality (lag) functions were considered in all capitals, presenting the occurrence of different lags according to the capital and the variable. Thus, the highest temperatures among the capitals analyzed occurred in Teresina/PI and the least high, in Salvador/BA. In general terms, the optimistic scenarios (C1) presented temperature between 32.5 and 35 ºC, the pessimists (C2) between 37.5 ºC and extremes (C3) 35 and 39 ºC, evidencing that all future scenarios present danger to the population. It is expected that the results obtained can help public policies.