Theoretical methods and algorithms have made great strides in their capacity to faithfully describe complex molecular systems, while, in parallel, there has been an enormous increase in the capability of and access to high performance computing resources. As a result, there is now a realistic expectation that theory and experiment should function in a symbiotic relationship to resolve the molecular mechanisms underlying interesting laboratory phenomena, and ideally point to experimental refinements in, e.g., the molecular composition of a material. A key challenge is that the molecular structures present are often unknown, and so one first must postulate one or more “molecular scenarios”, including structures and general mechanisms, and then use calculations to determine whether such a scenario leads to observables consistent with experiment. In this talk, I will discuss two such examples. The first is an observation of variations in the fluorescence quantum yield of a mutant GFP when linked to an active neuron; the second is an exploration of the determinants of molecular morphology during spin-coating of a thin-film polymer semiconductor. In each case, the results lead, much as we learn in chemical kinetics, to a mechanism that is consistent with experiment, but awaits further experimental tests.