With the alarming rise of antimicrobial resistance, studies on bacteria-surface interactions are both relevant and timely. Scanning electron microscopy and colony forming unit counting are commonly used techniques but require sophisticated sample preparation and long incubation time. Here, we present a direct method based on molecular dynamics simulation of nanostructured surfaces providing in silico predictions, complemented with time-lapse fluorescence imaging to study live interactions of bacteria at the membrane-substrate level. We evaluate its effectiveness in predicting and statistically analyzing the temporal evolution and spatial distribution of prototypical bacteria with costained nucleoids and membranes (E. coli) on surfaces with nanopillars. We observed cell reorientation, clustering, membrane damage, growth inhibition, and in the extreme case of hydrocarbon-coated nanopillars, this was followed by cell disappearance, validating the obtained simulation results. Contrary to commonly used experimental methods, microscopy data are fast processed, in less than 1 h. In particular, the bactericidal effects can be straightforwardly detected and correlated with surface morphology and/or wettability.