Optical sensing is an inexpensive and simple technique in water pollution control that offers high sample throughput and field applications. In this work, we applied an innovative reaction-based optical fingerprinting strategy for the recognition of clean and synthetically contaminated waters. Samples of spring, well, borehole, pond, and tap water were introduced into the reaction mixtures containing carbocyanine dyes, the absorbance and fluorescence of which were monitored photographically at different times; the data were processed by linear discriminant analysis. Up to 11 clean water samples were completely discriminated; the samples artificially polluted with the soluble fraction of different brand engine oil [water-soluble fraction (WSF)] and iron(2+) supplied as Mohr's salt were discriminated from uncontaminated ones; the presence of 35-fold diluted WSF in water was also detected. Iron(2+) was detected in water samples at 3 μg/L using the same indicator reactions. The proposed strategy can be useful in water pollution monitoring.