The effect of 16 liquid solvents on both the spectrum and molar absorption coefficient of the X-3 Sigma(-)(g) -> b(1)Sigma(+)(g) transition in molecular oxygen has been examined. The ability to monitor this weak transition using air or oxygen saturated samples at atmospheric pressure was facilitated by the rapid and efficient O-2(b(1)Sigma(+)(g)) -> O-2(a(1)Delta(g)) transition, which allowed the use of O-2(a(1)Delta(g)) phosphorescence as a sensitive probe of O-2(b(1)Sigma(+)(g)) production. The results of these O-2(a(1)Delta(g)) phosphorescence experiments are consistent with the results of independent experiments in which the O-2(a(1)Delta(g)) thus produced was "trapped" via a chemical reaction. The data recorded were used to calculate rate constants for the O-2(b(1)Sigma(+)(g)) -> O-2(X-3 Sigma(-)(g)) radiative transition, a parameter that is otherwise difficult to directly obtain from such a wide range of solvents using O-2(b(1)Sigma(+)(g)) -> O-2(X-3 Sigma(-)(g)) phosphorescence. The data show that the response of the O-2(b(1)Sigma(+)(g)) -> O-2(X-3 Sigma(-)(g)) radiative transition to solvent is not the same as that of the O-2(b(1)Sigma(+)(g)) -> O-2(a(1)Delta(g)) and O-2(a(1)Delta(g)) -> O-2(X-3 Sigma(-)(g)) radiative transitions, both of which have been extensively examined over the years. However, our data are consistent with a theoretical model proposed by Minaev for the effect of solvent on radiative transitions in oxygen and, as such, arguably provide one of the final chapters in describing a system that has challenged the scientific community for years.