Following an introduction into solvent radical-cage effects and the general use of cage-trapping methods to study solvent-cage effects, the bio-organometallic enzyme cofactor coenzyme B-12 (adocobalamin; AdoCbl) is studied using the TEMPO (2,2,6,6-tetramethylpiperidinyl-1-oxy) nitroxide cage-trapping method. Specifically, the products, kinetics, and mechanism of AdoCbl Co-C thermolysis in ethylene glycol are presented at high [TEMPO], results which establish (after ruling out other conceivable mechanisms and interpretations) a radical-cage effect for AdoCbl in ethylene glycol. The results also allow a limit to be placed on the important fractional cage efficiency factor (F(c)) for AdoCbl, 0.4 less-than-or-equal-to F(c) less-than-or-equal-to 1.0 (where F(c) is defined as the ratio of cage recombination to the sum of all competing cage processes, F(c) = k-1/SIGMAk(cage)). The findings suggest possibly important and more general, but little recognized, biological analogs of cage effects in coenzyme B-12-dependent enzymes. Also discussed is the quite speculative but novel and especially intriguing idea that B-12-dependent enzymes may include protein ''ultimate radical cage'' and ''ultimate radical trap'' effects.