Lakes play an important role in the global carbon cycle. Hence, data on seasonal and spatial variations of phytoplankton primary production (PP) is needed to estimate their ecological state. However, considerable difficulties arise when investigating the spatial and temporal variations of productivity in inland waters. Traditional in situ sampling approaches cannot provide the necessary data over large lakes or for lakes over large geographic regions. A solution can be to combine modelling with satellite data, by using different satellite products as inputs for a primary production model. The objective of this study is to evaluate and validate a lake primary production model based on satellite products. For that purpose we compared PP(sat) with PP(in situ) in two large, shallow, well mixed, and eutrophic Estonian lakes. A semi-empirical phytoplankton primary production model, derived by us earlier, was used. It allows estimating the values of phytoplankton primary production for lakes over large geographic areas. The primary production model operated well using MERIS products as input data. Our results show that spatial variability in primary production cannot be neglected in large lakes. Combining satellite data with model calculations allows monitoring of the state of lakes in space and time, providing a new perspective for lake primary production studies.