In our experience electrofishing in Greek, Cypriot and Turkish rivers, we find many very-polluted sites that are fishless; yet it seems difficult to use fish as indicators of specific pollution loads. Degradation is so multivariate. In building fish-based multimetric indices for ecological integrity assessment in rivers little is heard of tolerance to pollution. I am refering to Mediterranean-climate, low diversity, wide niche-breadth assemblages. Any refs would be appreciated.