http://dx.doi.org/10.5902/2179460X16207The homogenization is a procedure to improve the quality of data. In climatology, homogenization examines and adjusts temporal biases of climatic variables, caused by non-climatic factors. The objective of this study was to homogenize the data mean air temperature (Tma) applying the method ACMANT (Adapted Caussinus-Mestre Algorithm for homogenising Networks of Temperature series) and study trends of Tma in Aracaju-SE. ACMANT is a fully automatic in Fortran code that uses multiple relative time series for homogenization. The ACMANT is a further development of the Caussinus-Mestre method, that is one of the most effective tool among the known homogenising methods. The methodology of this study used ACMANTv1.2 version and monthly data of Tma for the period 1961-2012 to capital of Sergipe (Aracaju) and eight nearby cities obtained in Meteorological Database for Education and Research (BDMEP), the Climate Standard of the National Institute of Meteorology of Brazil (INMET) and application and Mann-Kendall test for non-parametric analysis of trend in the study of Tma series. In the series of original data existed 23.7% missing value that were filled by the method. The results the Mann-Kendall test, after the homogenization process, showed trends significant (p-value<5%) of elevated to Tma for the period with an increase of 0.5°C over the last decade (2000-2010) in reference to Climate Standard INMET of 26°C.A homogeneização é um processo para aperfeiçoar a qualidade dos dados. Em climatologia, a homogeneização analisa e ajusta tendências temporais de variáveis climáticas, causadas por fatores não climáticos. O objetivo deste trabalho foi realizar a homogeneização de dados da Temperatura média do ar (Tma) aplicando o método ACMANT (Adapted Caussinus-Mestre Algorithm for homogenising Networks of Temperature series) e analisar tendências da Tma em Aracaju-SE. O método ACMANT é totalmente automático em código Fortran e utiliza séries temporais em relação múltipla para o processo de homogeneização. O ACMANT é um aperfeiçoamento do método Caussinus-Mestre, que consiste em uma das ferramentas mais eficazes entre os métodos de homogeneização conhecidos.