Natural fluorescence as a method to monitor biofilm processes was studied, using the example of an extractive membrane bioreactor for the degradation of 3-chloro-4 methylaniline and 1,2-dichloroethane. Non-invasive, on-line, in-situ 2D fluorometry monitoring was employed to elicit biofilm process status. The fluorescence fingerprints were deconvoluted in a pattern recognition approach using artificial neural networks (ANN) through association with key process performance parameters.