Este trabalho teve por objetivo caracterizar as ondas de calor observadas no estado do Pará, região Norte do Brasil, durante 40 anos. Foram utilizados dados diários de temperatura máxima do ar para o período de janeiro de 1980 a dezembro de 2019 de 15 estações meteorológicas convencionais situadas na região de estudo. A Análise de Cluster foi empregada a fim de identificar regiões homogêneas de temperatura máxima do ar enquanto que a identificação das ondas de calor foi realizada usando os índices CTX90pct e GDE. Os resultados apontam por meio da Análise de Cluster que existem 3 regiões com comportamento homogêneo da temperatura máxima do ar, sendo a maior variabilidade na escala interanual, com destaque para os anos de 1998, 2005, 2010, 2015, 2016 que apresentaram o maior número de eventos registrados. A análise permitiu verificar que as ondas de calor apresentam similaridades espaço-temporal, com elevada variabilidade interanual e sazonal, informações que podem servir de subsídio para implementação de medidas públicas que visem reduzir os impactos negativos que as ondas de calor podem causar na população paraense.Palavras-Chave: Amazônia; desastres naturais; eventos climáticos extremos Climatology of Heat Waves in The State of Pará, Brazil A B S T R A C THeat Waves (HW) are one of the extreme climatic events that most impact the various branches of society. This study aimed to characterize the HW observed in the state of Pará, northern Brazil, for 40 years. Daily data of maximum air temperature were used for the period from January 1980 to December 2019, from 15 conventional meteorological stations located in the study region. Cluster Analysis was used to identify homogeneous regions of maximum air temperature, while the identification of HW was performed using the CTX90pct and DD indices on an annual, seasonal and monthly scale. Six characteristics of HW were evaluated: occurrence, frequency, mean duration, mean intensity, maximum intensity and severity degree, so that through student's t-test it was verified whether the means of the characteristics of the HW are equal between the clusters. The results indicate through CA that there are 3 regions with homogeneous behavior of maximum air temperature, being the greatest variability in the interannual scale, especially for the years 1998, 2005, 2010, 2015, 2016 that presented the highest number of events recorded. The t-test showed that in clusters, for the six characteristics, a similarity in the mean behavior is observed, indicating that the events of HW present spatial similarity in the state. The analysis allowed us to verify that HW present space-temporal similarities, with high interannual and seasonal variability, information that can serve as a subsidy for the implementation of public measures aimed at reducing the negative impacts that heat waves can cause on the population of Pará state.Keywords: Amazon; natural disasters; extreme weather events