A　new　self-organizing　algorithm　for　fuzzy　neural　networks　is　proposed,　which　automates　the　structure　and　parameter　identification　simultaneously　based　on　input-target　samples.　Firstly,　a　self-organizing　clustering　method　is　used　to　establish　the　network　structure　and　the　initial　values　of　its　parameters.　Then　a　supervised　learning　is　applied　to　optimize　these　parameters.　An　example　of　nonlinear　function　approximation　is　given　to　demonstrate　the　effectiveness　of　the　algorithm,　where　some　comparisons　are　made　with　other　approaches.　Finally,　based　on　the　data　of　a　wastewater　treatment　plant,　a　forecast　model　of　the　output-water　quality　is　developed　using　the　established　fuzzy　neural　networks.　Simulation　results　show　that　the　output-water　quality　can　be　well　predicted　by　the　model.