The statistics of selenium chain length distribution in GexSe100-x glasses with 5 <= x <= 20 are investigated using a combination of high-resolution, two-dimensional Se-77 nuclear magnetic resonance (NMR) spectroscopy and quantum chemical calculations. This combined approach allows for the distinction of various selenium chain environments on the basis of subtle but systematic effects of next-nearest neighbors of Se atoms in -Se-Se-Se- linkages on the Se-77 chemical shift tensor parameters. Simulation of the experimental Se-77 NMR spectral line shapes indicates that Se chain speciation in these chalcogenide glasses follows the Flory-Schulz distribution, originally developed for organic chain polymers.