The origin of the long wavelength fluorescence emission (peak-to-peak Stokes shifts similar to 9050-9450 cm(-1)) of 3,5-di-tert-butylsalicylic acid (t-BSA) has been investigated in a variety of organic solvents. Two types of emissions, I and II, can be identified. The type I emission is observed in hydrocarbon and nonpolar solvents. It is a relatively weak fluorescence (phi(f) less than or equal to 0.01), and the lambda(F) lies at similar to 466 nm. The type II emission is relatively strong, and phi(f) values ranging from 0.2 to 0.6 are obtained. It is observed in alcohols, acetate solvents, and acetonitrile. The lambda(F) is at similar to 430 nm. Mixed solvent experiments suggest that solvent molecules H-bond with t-BSA to form cluster complexes in these solvents and that these cluster complexes emit to give rise to the type II emission. While the fluorescence intensities of both types of emissions are quenched by acetic acid, the introduction of a trace amount of triethylamine is shown to enhance the type I emission (in cyclohexane) but not the type II emission (in ethanol). This observation, along with the similar fluorescence spectra between t-BSA and LitBSA (lithium 3,5-di-tert-butylsalicylate), suggests that the long wavelength emission of t-B SA originates either from the excited 3,5-di-tert-butylsalicylate anion or a species derived from it.