This paper is the first in a series of two articles where we report the development of fast sugar structure prediction software (FSPS). To the best of our knowledge, this is the first automated tool for the systematic study of conformations of complex oligosaccharides in solution. In contrast to previously developed molecular builders such as POLYS (Engelsen, S. B.; Cros, S.; Mackie, W.; Perez, S. Biopolymers 1996, 39, 417-433) where only information about the minimum energy conformation of disaccharide pairs is considered in order to build larger oligosaccharides, this tool is based on a systematic search of dihedral conformational space, optimization of structures using implicit solvation models, explicit molecular dynamics simulations, NOE calculations, and a very powerful substructure recognition algorithm and database. Our FSPS can rapidly find minimum-energy conformers and rank them according to different criteria. Two such criteria are the energy of the conformers in implicit solvent and the root-mean-square deviation (RMSD) of computed NOEs with respect to experimental data. Even though experimental NOEs may result from an average over conformers instead of a single structure, we find that sorting according to NOE RMSD constitutes a better estimator for the global free-energy minimum structure in explicit solvent (i.e., the most likely structure in solution). In contrast, the lowest-energy structure in implicit solvent does not usually correspond to the free-energy minimum. A harmonic approximation to compute free energies of each conformer does not appear to reverse this conclusion, indicating that either explicit hydrogen bonding to the solvent or anharmonic effects in the free energy or both are fundamentally important. In the first article, we discuss our methodology and study, as a proof of concept, a simple substituted disaccharide. In the second article, we focus on two complex human milk oligosaccharides.