The infra-red and Raman spectra of cyclopentane, cyclopentane-d1, and cyclopentane-d10 have been determined for the purpose of establishing the symmetry of cyclopentane. D5h selection rules are found to hold very well. This does not constitute a rigorous criterion for structure in this case, however, because drastic alteration of the symmetry by substituting groups such as D, OH, CH3, or Cl for a hydrogen atom does not appreciably increase the complexity of the spectra. The data have provided two strong arguments against a D5h structure. (1) An assignment could not be made which simultaneously satisfied the product rule and the expected band contours. (2) The entropy of the vapor demands a low frequency (near 140 cm−1 for the assignment in this paper) if the symmetry is D5h. This is completely incompatible with the heat capacity of the solid. It is concluded that cyclopentane definitely does not have D5h symmetry. The actual geometry of the molecule is still not known. It is shown that a rigid structure of symmetry Cs, C2, or C1 is consistent with the available data. A decision between this and Kilpatrick, Pitzer, and Spitzer's model with a puckered ring and a pseudorotation in place of one of the genuine normal vibrations cannot be made at this time. Analogous spectroscopic results for perfluorocyclopentane are mentioned briefly for comparison. This molecule is definitely non-planar. The method of preparation of the deuterium compounds is given. A Pd-on-charcoal catalyst is described which promises to be useful in other hydrogen-deuterium exchange reactions.