The question of the structure of the second stable conformer of butadiene (cis or gauche) has been reexamined by using a combination of theoretical and experimental methods. High level MP3/6-311+G**//MP2/6-31G* ab initio calculations predicted the gauche conformer to be energetically preferred by 0.98 kcal/mol (0.85 kcal/mol after correction for zero-point energies) over the cis conformer. Experimental data were reexamined as follows. An ab initio derived force field for s-trans-butadiene was fit to the observed gas-phase infrared and Raman spectra. Infrared intensities for s-rrans-butadiene were measured and converted to dipole moment derivatives with respect to the internal coordinates, and the derivatives were compared to those obtained theoretically. The intensity data proved useful in determining the form of the normal coordinates for the out-of-plane bending modes.These data should prove useful in comparisons with force fields and dipole moment derivatives for other alkenes. The scaling factors obtained in the normal coordinate analysis were transferred to the calculated force fields for the cis and gauche forms, and the vibrational spectra were derived from these data. The ratios of dipole moment derivatives between experiment and theory, along with the calculated derivatives for the other rotamers, were used to predict intensities for both cis- and gcucAe-butadiene. The constructed spectra for the minor rotamers were compared to the experimental spectra. gauche-Butadiene was found to fit the data better than the cis conformer, in agreement with ab initio calculations. Vertical transition energies for the π to π* (Ag → Bu) transition for cis, gauche, and trans rotamers were calculated. While absolute transition energies were 5-10% too large, relative energies (with an origin corrected to trans-butadiene) supported a gauche conformer -5to +10 nm to the red of the trans form and a cis form nearly 30 nm to the red. Although the second conformer was previously reported to be 14 nm to the red of the trans form, new experimental data suggested this gap may be only 3 nm. Significantly, both assignments of »max were in agreement with a twist angle near 25-35°. This correlation of »max to twist angle was supported empirically with the known data for cycloheptadiene and cyclooctadiene. © 1990, American Chemical Society. All rights reserved.