It is a profound belief of biologists, based upon decades of structural biochemistry, that understanding the three-dimensional structure of individual proteins and protein complexes illuminates the basis of protein function. X-ray crystallography and other structural techniques have revealed remarkable high-resolution maps of even large, multimolecular structures. However, the approaches which give such spectacular information in vitro are unable to probe structures of proteins that are functioning in their natural milieu in living cells. This frontier between structural biology and cellular biology is only beginning to be addressed. An example of the progress is work from the laboratory of Sanford Simon (Rockefeller University, New York). Earlier works by Mattheyses et al. (1) and Kampman et al. (2), and a new article by Atkinson et al. (3) (this issue) describe the elegant application of fluorescence polarization to probe the orientation of specific protein components in functioning nuclear pore complexes in living cells. Nuclear pores are an essential component of eukaryotic cells that are responsible for the highly regulated, bidirectional transport of proteins and RNA between the cytoplasm and nucleus. They are very large structures that contain a large aqueous channel that spans the double membrane of the nuclear envelope. Pores are composed of ∼30 different proteins (nucleoporins) present in multiple copies, with several thousand proteins comprising a single nuclear pore. Crystallography has provided structural information about isolated nucleoporins and their interaction sites (4). Electron microscopy of intact nuclear pore complexes has revealed a structure with octagonal symmetry (5,6). Simon and colleagues have taken advantage of the random distribution of nuclear pores around the approximately circular cross section of the nuclear envelope. Nuclear pore proteins labeled with green fluorescent protein were introduced into cells by molecular biology techniques. By illumination of the sample with polarized light and collection of the emission in two orthogonally polarized channels, the fluorescence anisotropies of nuclear pores with their varying orientations around the nuclear envelope cross section were determined. Thus, the intrinsic geometry of the positions of the nuclear pores in the nuclear envelope enabled investigation of the anisotropy at various angles. Channel nucleoporins have a predicted α-helical region and an unfolded domain containing phenylalaine-glycine (FG) repeats. The ensemble of FG domains from multiple nucleoporins in a channel is thought to function as a diffusion barrier to regulate transport of large macromolecules (Fig. 1 (7)). This article addresses the question of whether FG regions in different nucleoporins are oriented relative to the central axis of functioning nuclear pores in living cells. The findings were significant: some of the FG domains of specific nucleoporins were ordered. Ordering was more likely for nucleoporins lining the center of the nuclear pore complex than for those at the cytoplasmic or nuclear entrances. Ordering was not altered by active transport or attachment of cargo. Although one might have predicted (wished?) a change with function, the time resolution of anisotropy measurements (2 s) was likely insufficient to capture fluctuations due to individual transport events. Figure 1 Schematic of nuclear pore. Adapted from Fig. 12-10 in Alberts et al. (7). These studies are a hopeful beginning, demonstrating the possibility of asking structural questions within living cells. We can look forward to improved techniques, including faster time resolution approaches and the application of single-molecule fluorescence imaging, that will enable the detection of function-related structural changes in proteins in their normal, intracellular milieu. These types of biophysical techniques should be applicable to numerous other dynamic protein structures within cells.