The southeastern Brazilian coast is a vulnerable region to the development of severe storms, mainly caused by the passage of cold fronts and extratropical cyclones. In the last decades, there has been an increase in the occurrence of subtropical cyclones. This study investigates trends and climatic variations, analyzing surface meteoceanographic series at six grid points from the reanalysis databases of ERA-Interim and ERA5 (European Center for Medium-Range Weather Forecasts (ECMWF)) from 1979 to 2018 over the ocean region bounded, approximately, at 18°S, 25°S, and 37ºW, 45ºW (between the states of Espírito Santo, Rio de Janeiro, and São Paulo). Non-parametric statistical tests and the generalized extreme value distribution are employed for annual, seasonal, and daily maxima/minima. The numbers of occurrence of extreme values, as well as the extremal index, are also estimated in order to better understand the behavior of extremes. Annual maximum sea-surface temperature anomalies of the ERA-Interim databases show very low negative values, mainly at the beginning of measurements (between 1979 and 1982), leading to high positive trend values. The results are compared to the updated data from ERA5 which have anomalies that are more homogeneous with positive trends but without statistical significance. The other meteorological series of the ERA-Interim does not present discrepancies. Only the maximum anomalies of air temperature have significant annual and seasonal positive trends at grid points near the coast of Rio de Janeiro and São Paulo. Despite that the analyses for pressure and wind speed anomalies do not indicate significant trends, they present increases in the interdecadal pattern of the numbers of occurrence of extreme percentiles for almost every grid point. Return levels for 10, 25, 50, 75, and 100 years are estimated at each grid point, and many maximum/minimum peaks are close to the return levels for 100-year return periods. The extremal index suggests average cluster sizes associated with no predominance of clustering for the extreme percentiles, which represents weak dependence between the exceedances. These results characterize some independence between extreme meteorological events such as the event that has been taking place in the region. The occurrence of maximum daily wind speed peaks calculated in austral spring, whose values exceeded the previous ones, is identified at three grid points near the southeast Brazilian coast, caused by the passage of the subtropical cyclone “Deni,” which occurred in November 2016.