In this paper, we propose the proof of principle of a low-cost, autonomous battery-powered MEMS-based platform bearing molecularly imprinted polymer dots for the parallel recognition of specific target molecules. Though it is taken for granted that medical agencies in developed countries are erratically concerned with possible pandemic like those arising from the latest strain H5N1 influenza or West Nile virus, people in developing regions deal with the repercussions of pandemics on a daily basis (figure 1). If no doubt exists concerning technological advances should benefit to many of these regions, everyone is aware with the fact that current diagnostic products are often developed with the major Western markets in mind [1]. Indeed, reagent storage, power supply, sample preparation and other issues are all impacted by factors such as geographic location. Thus, no electrical power is available in the most remote rural areas, neither conventional storage can be applied to fragile biological reagents in tropical regions as mostly in case of developing-world countries. Sometimes the solution to these issues can be as easy as removing the need for an external source of electricity and proving new ways of specifically functionalizing sensors with stable, ruggedized biorecognition layers. For this to be achieved, the combination of piezoelectric MEMS resonators and molecularly imprinted polymers (MIP) reaches the aforementioned specifications (figure 2). Matrices of circular micromembranes with a piezoelectric patch for an integrated excitation-detection scheme were fabricated and used with a home-made electronic set-up allowing the precise determination of the resonant frequencies of the devices (figure 3). The dedicated electronic was developed for a 12 DC volts battery alimentation and contains a dynamic compensation of electrical parasitic elements, permitting the multiplexed and precise measurement of resonance spectra [2]. After an individual functionalisation of the structures by MIP thanks to a cantilever-array based deposition tool [3], the multiplexed format of the chips showed the potentiality of the combined MEMS and MIP system for the specific detection of target molecules. In the present work, herbicide 2,4-Dichlorophenoxyacetic acid (2,4-D) was chosen as a template molecule for its well-known behavior when imprinted in polymers. Figure 4 shows the reproducibility of frequency measurements after several templates washing and incubation cycles. As resonant frequency is sensitive to mass variations on the sensor's surface, the withdrawal of the 2,4-D molecules resulted in an increase of the frequency (since mass decreased) while after rebinding of the templates in aqueous buffer, it decreased with a signal-to-noise ratio of 0.3%. Reproducibility of the measurements showed the potentiality of the combined system as a reusable chip, significantly decreasing experiment costs. Validity of the label-free detection principle was confirmed by the negligible shifts of the frequency in the case of a micromembrane bearing a non-imprinted polymer (used as a control). Moreover, such template recognition was specific since a 10 times higher concentration was necessary for the same frequency levels to be reached when polymer was incubated in phenoxyacetic acid (POAc), a similar chemically structured molecule and this result completely fits with the literature (figure 5) [4]. Sensitive detection was also proven since a 10µM minimum detectable concentration was obtained in this label-free, multi-sensor approach. These results correspond to the first combination of MEMS and molecularly imprinted polymers in a suitable format for the quantitative detection of small molecules. Work is now under progress for the study of several molecularly imprinted polymers integrated on matrices of piezoelectric micromembranes. Main objectives deal with the parallel and quantitative detection in liquid and in real-time of the templates molecules, completing the system for a portable, autonomous, label-free and low-cost biological analyses platform.