There is evidence that human telomerase acts as a dimer [Wenz, C., Enenkel, B., Amacker, M., Kelleher, C., Damm, K., Lingner, J., 2001. Human telomerase contains two cooperating telomerase RNA molecules. EMBO J. 20, 3526–3534]. Three possible mechanisms have been proposed. We translate those proposals into three detailed mechanistic models for telomerase action, also introducing optional isomerizations with equilibrium constants inversely related to the number of bound nucleotides. To distinguish between these models by in situ experiments we propose a microscopic system which uses two-photon excitation of fluorescence in a volume of about 0.5 μm3. A variety of detection strategies and experimental designs are considered; we focus on those best suited to observation of a small volume under limitations imposed by diffusion to and from the reacting micro-volume, and consequently restrict ourselves to constant flow. Numerical simulation is used to help identify an optimal experimental design. The detection of mechanistic changes hinges on linking fluorescence reporters to selected reaction components, either directly (chemically) or indirectly (via an indicator reaction). We show that rapid mixing experiments are better than chemical relaxation experiments, as the statistics of single molecule kinetics affects the latter more than the former. However, some fast reaction steps can only be revealed by chemical relaxation coupled with mixing experiments. We explore connections between our methods and studies of HIV and other systems with RNA to DNA transcription.