FFS is an attractive candidate for studying protein interactions in cells and is based on fluctuations in the fluorescence intensity observed in a small observation volume (Ͻ1 fl) that are due to individual proteins entering or leaving the volume (1). Fluorescence correlation spectroscopy (FCS) uses correlation functions to determine the concentration and temporal properties of proteins, and dual-color FCS has been used to detect protein interactions in cells (2, 3). However, a quantitative characterization of protein interactions has proven to be difficult. We argued previously that brightness analysis of fluctuations is a promising method for quantifying protein interactions in living cells (4). The brightness of a molecule is defined as the average fluorescence intensity of a single particle, and analysis methods that determine the brightness from fluctuation data are available (5, 6).Brightness encodes the stoichiometry of protein complexes. Consider the case of a dimer. If the monomeric protein carries a fluorescent protein of brightness then the homodimer exhibits a brightness of 2 , because it carries two fluorophores. This concept was experimentally verified by using GFP as a marker and applied to study the concentration-dependent dimerization of nuclear receptors (NRs) in living cells (7). However, the use of a single fluorescent color limits brightness analysis to homocomplex formation. Two fluorescent colors are required to tackle heterointeractions. We developed the analysis tools for brightness analysis of two colors and tested it by using the heterodimer of the ligand-binding domains (LBD) of the NRs retinoic X receptor (RXR) and retinoic acid receptor (RAR) in CV-1 cells (8). Quantifying heterointeractions is considerably more complex than the analysis of homocomplex formation, and a method that identifies a heterocomplex of unknown composition is currently unavailable. In this manuscript we devise a methodology that solves this problem and identifies the stoichiometry of two proteins labeled with CFP and YFP within a cellular protein complex. We also introduce the brightness profile, which characterizes the normalized brightness as a function of the coexpression ratio of the labeled proteins. The brightness profile is used together with brightness titration experiments to deduce the composition of a protein complex at stoichiometric binding conditions. We illustrate the method using a protein complex that is larger than a dimer to highlight the inherent strength of brightness analysis.NRs such as RXR are known to interact with coactivators. It is generally thought that two NRs bind a single coactivator molecule, thus providing a suitable test system that goes beyond the simple dimer model. We choose the LBD of RXR and the NR interacting domain (NID) of SRC1 (steroid receptor coactivator 1) as a minimal model for exploring the formation of a hetero-oligomer by FFS directly in living cells. Brightness analysis revealed that RXRLBD and SRC1NID form a heterotetramer with three RXRLBDs binding to one ...