Confocal and two-photon fluorescence microscopy techniques using genetically encoded fluorescent probes are widely used in cell biology. Beyond the common problems of photobleaching and phototoxicity, we present evidence that photounbinding also has the potential to compromise such methods, especially in quantitative studies. We show that laser intensities within excitation regimes typical for imaging approaches such as as fluorescence recovery after photobleaching (FRAP), photolysis, or fluorescence correlation spectroscopy (FCS) experiments can cause the dissociation of antibodies from their ligands. Indeed, both one- and two-photon excitation of a fluorescent anti-GFP antibody caused its dissociation from immobilized GFP in vitro. Importantly, with two-photon excitation, the laser intensity threshold for photobleaching was the same as for photounbinding. By contrast, with single-photon excitation, we found a range of laser intensities where photobleaching can be separated from photounbinding. This photounbinding effect was visualized and measured by rebinding a second fluorescent anti-GFP (Green Fluorescent Protein) antibody, indicating that the GFP remained functional for reassociation following the photoinduced dissociation. Finally, we show that this unbinding effect occurs only when at least one binding partner carries a fluorescent label. Our results show that this photounbinding effect can readily remain masked or be misinterpreted as photobleaching, which can compromise the quantitative interpretation of binding studies made using fluorescence microscopy.