Within the last two to three decades, molecular imprinting has established itself a valuable technique for generating biomimetic recognition materials for a wide range of molecular and supramolecular analytes [1]. Among different target species, those with dimensions at the nanoscale are of special interest, because usually no easy-to-operate rapid tests or sensors exist for that size range. However, whereas there are some MIP for biogenic “nanoparticles”, i.e proteins [2] or viruses [3], this is not the case for engineered nanoparticles (ENPs), but for one exception: The group of D. Mandler [4] reported on a MIP based on electropolymerization to bind Au NPs. Despite successful imprinting, electropolymerized systems offer somewhat limited choice for finding optimal functional groups to interact with the respective particles. Polyadditions and radical polymerizations allow for much higher variability: We succeeded in surface imprinting of both polyurethanes and self-initiating polyacrylates, respectively, using nanoparticles as templates, as can be seen in the AFM image shown in Figure 1: It clearly reveals cavities of the desired size in the surface of a polyurethane MIP. Coating such polymers as recognition layers on quartz crystal microbalances (QCM) yields concentration-dependent frequency signals for Au NPs. One can clearly see that the MIP takes up almost ten times more NPs, than the corresponding non-imprinted polymer (NIP). Overall, such sensors hence allow for detecting Au NPs down to concentrations of a few ppm in aqueous solution with a dynamic range spanning at least 1-2 orders of magnitude in concentration. The layers efficiently exclude larger particles, as expected during imprinting. Preliminary tests reveal that ENP QCM sensors show selective behavior depending on particle diameter, particle composition and stabilizer. Furthermore, the concrete imprinting strategy has substantial influence on the sensor results: Using the same Polyurethane recipe, we developed three different imprinting protocols for ENPs, namely one-step imprinting, stamp imprinting, and sedimentation imprinting. One-step imprinting comprises mixing all monomers and templates in one batch and coating the respective sensor surface with the resulting pre-polymer. Stamp imprinting utilizes stamps containing (covalently bound) NPs to generate surface structures. Sedimentation imprinting means overlaying a spin-coated pre-polymer with the template NPs before hardening. One-step imprints did not yield any difference in sensor signal between MIP and non-imprinted polymers (NIP). Of the other two, sedimentation imprinting leads to highest overall signals (on both MIP and NIP). However, stamp imprinting leads to the largest imprinting factors: on average, MIP give rise to signals that are almost ten times larger, than those of corresponding NIP. PU-based ENP MIP have hence led to appreciable sensor responses. However, there is room to improve batch-to-batch reproducibility, because highly cross-linked polymers resulting from polyaddition turned out inhomogeneous and bearing rough surfaces. First results with self-initiating acrylates to overcome this are promising: Such MIP are indeed useful for incorporating ENPs in a reversible and quantitative manner, though imprinting effects at the current stage are lower, than in case of polyurethanes: MIP signals are usually 2-3 times higher, than NIP signals. In this light, several questions still need answers: To name only the most important ones, the exact recognition mechanisms are yet unknown; what is the exact influence of the NP core and the stabilizer shell, respectively, on the sensor signal? This notwithstanding, MIP have turned out a highly promising tool for selectively binding artificial nanoparticles from aqueous samples, which makes them interesting for e.g. food and cosmetics analysis. ACKNOWLEDGEMENTS This work was funded partly by the EU Commission in the 7th Framework Program for Research, (grant no. 280550 “INSTANT”) and partly by the Austrian Science Fund FWF, (project No. I 3568-N28), which we gratefully acknowledge. FIGURE CAPTION: AFM image and QCM sensor response, respectively, of Au-NP surface MIP. REFERENCES Ye, K. Haupt, Molecularly imprinted polymers as antibody and receptor mimics for assays, sensors and drug discovery. Anal. Bioanal. Chem. 378 (2004) 1887-1897. Naklua, R. Suedee, P. A. Lieberzeit, Dopaminergic receptor–ligand binding assays based on molecularly imprinted polymers on quartz crystal microbalance sensors. Biosens. Bioelectron. 81 (2016) 117-124. Wangchareansak, A. Thitithanyanont, D. Chuekheaw, M. P. Gleeson, P. A. Lieberzeit, C. Sangma, Influenza A virus molecularly imprinted polymers and their application in virus sub-type classification. J. Mater. Chem. B 1 (2013) 2190-2197. Kraus-Ophir, J. Witt, G. Wittstock, D. Mandler, Nanoparticle-imprinted polymers for size-selective recognition of nanoparticles. Angew. Chem. Intl. Ed. 53 (2014) 294-298. Figure 1