Uruguay's freshwater network is threatened by widespread Harmful Algal Blooms (HABs) known to be triggered by human-related stressors such as land-use change and urban/industrial effluents. Existing field-based monitoring practices are limited due to their sparse spatial and temporal coverage. A complementary approach these techniques is to utilize remotely sensed observations for estimating optically relevant water-quality (WQ) parameters, of which chlorophyll-a (Chla) is a robust proxy for HAB quantification. There is, however, a lack of information on apt country-scale image processing schemes, i.e., the best combination of Chla algorithms, atmospheric correction (AC) methods, and satellite sensors that agree best with in situ Chla across Uruguay’s inland and coastal waters. Here, we analyze the accuracy of three different combinations of ACs (SeaDAS, POLYMER, and ACOLITE) and 17 Chla models applied to the Operational Land Imager (OLI), Multispectral Instrument (MSI), and Ocean and Land Color Imager (OLCI) onboard Landsat 8, Sentinel-2A/B, Sentinel-3A/B, respectively. The performance of different processing schemes was assessed both in terms of their numerical consistency with in situ Chla and classification accuracy for discriminating low vs. high Chla with an 8 mg m−3 decision boundary. Our results show that the Mixture Density Networks (MDN) algorithm is often among the top performers. Other strong results were achieved by Gons (2 bands), Moses (3 bands) and Normalized Difference Chla Index algorithms. Regarding the atmospheric correction processors, POLYMER works better for OLCI, and SeaDAS for the OLI, while no clear distinction among AC methods was found for MSI. Furthermore, the MDN model was also among the most reliable for assigning water pixels to low or high Chla ranges. This could represent a key criterion for discriminating water bodies with good ambient conditions critical for reporting nationwide Sustainable Development Goal (SDG) 6.3.2 and other monitoring applications.