Computational methods in chemistry speed up the discovery of bioactive compounds and make lab work cheaper and more effective. This starts with the automated mining and analyses of public data and boils down to molecular interaction modelling. Here we concentrate on the latter part exemplified by the development plant PARP-enzyme inhibitors as potential agrochemicals that enhance the survival and yield of crops under drought stress (cf. climate change). The same works for human drugs (e.g. human PARP is an anticancer target). Following the initial generation of a three-dimensional model of a protein target, millions of structures in different conformations are docked and filtered for interaction quality, availability and other properties in a virtual screening procedure. A resulting set of actual compounds is then evaluated in a new bioassay that does not require spraying of plants. It is the first test for drought stress tolerance that uses sterile plant clones in microtiter plates, is fully quantifiable, and concentration dependent. Sources: Tennstedt, S., Fischer, J. Brandt, W., Wessjohann, L. in: Virtual screening – tools for a faster selection of new drug leads”. Medicinal Chemistry in Drug Discovery – Review Book, 2013: 219-236 (Ed. Dubravko Jelić), Transworld Research Network (37/661 (2), Fort P.O., Trivandrum-695 023, Kerala, India), ISBN: 978-81-7895-560-5