The photoactive yellow protein (PYP) is a bacterial photosensor containing a para-coumaryl thioester chromophore that absorbs blue light, initiating a photocycle involving a series of conformational changes. Here, we present computational studies to resolve uncertainties and controversies concerning the correspondence between atomic structures and spectroscopic measurements on early photocycle intermediates. The initial nanoseconds of the PYP photocycle are examined using time-dependent density functional theory (TDDFT) to calculate the energy profiles for chromophore photoisomerization and proton transfer, and to calculate excitation energies to identify photocycle intermediates. The calculated potential energy surface for photoisomerization matches key, experimentally determined, spectral parameters. The calculated excitation energy of the photocycle intermediate cryogenically trapped in a crystal structure by Genick et al. [Genick, U. K.; Soltis, S. M.; Kuhn, P.; Canestrelli, 1. L.; Getzoff, E. D. Nature 1998, 392, 206-209] supports its assignment to the PYPB (1(0)) intermediate. Differences between the time-resolved room temperature (298 K) spectrum of the PYPB intermediate and its low temperature (77 K) absorbance are attributed to a predominantly deprotonated chromophore in the former and protonated chromophore in the latter. This contrasts with the widely held belief that chromophore protonation does not occur until after the PYPL (1(1) or pR) intermediate. The structure of the chromophore in the PYPL intermediate is determined computationally and shown to be deprotonated, in agreement with experiment. Calculations based on our PYPB and PYPL models lead to insights concerning the PYPBL intermediate, observed only at low temperature. The results suggest that the proton is more mobile between Glu46 and the chromophore than previously realized. The findings presented here provide an example of the insights that theoretical studies can contribute to a unified analysis of experimental structures and spectra.