The effect of the sample flow-rate, exposure time to the herbicide, biomass loading, biosensor film thickness, intensity of the actinic light, illumination cycle, and temperature on the biosensor response has been evaluated using waterborne simazine as test bench. The biosensing device is able to provide in situ measurements of the herbicide concentration every 180min. The biosensor limit of detection for this herbicide was 12?gL-1, with a working range of 50-800?gL-1. The biosensor specificity to simazine has been assessed by comparing its response to that of isoproturon. © 2013 Elsevier B.V.