In order to contribute to the effective fight against eutrophication of the Ebrié lagoon estuarine waters, the present study proposed a mathematical model to predict water transparency based on five physico-chemical descriptors (temperature, dissolved oxygen, nitric nitrogen N-NO 3 -, nitrous nitrogen N-NO 2 and ammoniacal nitrogen N-NH 4 +). On twenty-seven (27) water samples collected from 2014 to 2015, in situ measurements and physico-chemical analyses were carried out based on French standards. All the data collected was subjected to a multiple linear regression analysis. The proposed linear model is accredited with good statistical indicators. These statistical indicators revealed effective predictions with a coefficient of determination (R 2 = 0.9308) close to one; a standard deviation (RMSE = 0.1311) of less than 0.3; a cross-validation coefficient of determination (Q 2 CV = 0.9256) of more than 0.9; and a Fischer test p-value of less than 0.0001. Furthermore, the values of the ratio of theoretical to experimental transparency of the validation set tend towards one. The results obtained suggest that the combination of these 5 descriptors could be useful for predicting transparency. Moreover, dissolved oxygen is the priority descriptor for the prediction of the transparency of the Ebrié lagoon, although the influence of the other descriptors is not less important.