Fluorescence correlation spectroscopy (FCS) has been used to study the diffusion of lipid and protein molecules in both synthetic and natural membranes. Recently, a new version of FCS, inverse FCS, has been developed (Wennmalm et al., 2009). For inverse FCS fluorescent probes are present at very high concentration as opposed to low concentration for conventional FCS. The particle of interest in the inverse FCS experiment is non-fluorescent. Fluorescence fluctuations due to the exclusion of fluorescent probes from the excitation spot by the non-fluorescent particles are detected and correlated. In the study of membrane nanodomains by conventional FCS, both domains and probe diffusion contribute to the correlation fluctuation. However, for inverse FCS, the diffusion of fluorescent probes contributes little to the fluctuation due to the high concentration of fluorescent probes. Therefore, inverse FCS provides an opportunity to study the properties of dark nanodomains that exclude the fluorescent probe when the fraction of nanodomains in the membrane is small. Here, we demonstrate by simulation that in these situations inverse FCS can give information about membrane nanoscopic phase separation that is inaccessible to conventional FCS. We also used inverse FCS to experimentally study early phase separation in DLPC/DSPC model membranes. Two different domain evolution pathways have been observed. In one of these two pathways, nanoscopic domains appear at first and then gradually grow to micron size. In the other pathway, the domains reach micron size quickly and their number gradually increases.Inverse-Fluorescence Correlation Spectroscopy, Stefan Wennmalm, Per Thyberg, Lei Xu, and Jerker Widengren. Anal. Chem. 2009, 81, 9209-9215