We report results of semiempirical intermediate neglect of differential overlap (INDO/S) calculations on the spectroscopy of 4 -hydroxy-1-methylstilbazolium betaine (HMSBB). Solvent effects were included using both the continuum and supermolecule approaches. The calculated spectra are compared with experimental ultraviolet-visible spectra for several solvents. The odd bathochromic shift of the low energy band (for nonpolar solvents) is tentatively discussed in terms of the contributions of different solvation mechanisms, and an assignment based on INDO/S calculated spectra is presented. We start from the assumption that in solution, conformations other than the completely planar structure (A) are possible. For protic solvents, in which hydrogen bonding constitutes the main contribution to solvation, calculations using the planar, fully optimized AM1 geometry reproduce fairly well the observed spectra. The intense, π → π * band is calculated close to 20,000 cm −1 , in good agreement with the experimental values in water and methanol. For the very polar formamide, the INDO/S+SCRF result is also very accurate, suggesting that formamide's strong polarity induces HMSBB to adopt the planar structure A. For polar solvents with nonspecific interactions such as acetonitrile or dimethylformamide, the situation is not as clear. We speculate that the experimental numbers can be reproduced only by a very large supermolecule. Finally, for nonpolar solvents such as cyclohexane, benzene and CCl 4 , the experimental number of 16,141 cm −1 (previously obtained from extrapolation of solvent mixtures by Jacques and coworkers) is reproduced by the INDO/S method only considering structures B and C. These results can be considered a confirmation of Jacques' suggestion that the solvatochromism of HMSBB can be explained by the solvents inducing a geometry change in HMSBB.