Remote sensing data can give the spatial and temporal distribution of chlorophyll-a, which is impossible with field measurements. Chlorophyll-a can be considered crucial due to the fact that it characterizes the level of eutrophication of a marine system. The major aim of this paper is to assess the chlorophyll-a retrieval algorithms from satellite images using in situ estimations in the region of Southern Aegean Sea. A data set from the Copernicus Marine Environmental Service (CMES) containing in situ chlorophyll-a concentrations was used to evaluate ocean color retrieval algorithms. Images captured from the Sentinel-2 satellite were used. Methodologically, the images were atmospherically corrected, pixel clouds were removed, and the Maximum Band Ratio was calculated. Then the ocean color algorithm for the Mediterranean Sea (MedΟC3) was used to calculate the chlorophyll-a concentrations. The in situ data measurements of chlorophyll-a concentrations were obtained at a depth of 20, 50, 75 and 100 m. The hypothesis for a homogenous sea was used (temperature difference of ΔΤ<0.2°C) in order to assume that the concentration of chlorophyll-a is the same at the surface as in 20 m depth. Α fourth order polynomial equation was fitted to the observed data for estimating the error of retrieval algorithm. Also, linear regression models were utilized between reflectance of a single band, logarithmically transformed band ratios of the visible spectrum and in situ concentrations of chlorophyll-a. Scatter plots, histograms and statistical indexes were calculated in order to evaluate the results. The best fit was calculated using the fourth order polynomial relationship between in situ and satellite data. On the contrary, linear regression model were not able to estimate accurately the chlorophyll-a concentration.