The electrode reaction Co(diamsar)(3+/2+) in water at 25 degrees C, ionic strength I = 0.14 mol L(-1), is characterized by E degrees’ = -0.24(5) V and Delta V = +17.4 +/- 0.5 cm(3) mol(-1) vs Ag/AgCl/4 mol L(-1) KCl, and the kinetics of the corresponding self-exchange reaction in homogeneous solution by k(ex) = 0.48 +/- 0.01 L mol(-1) s(-1), Delta H-ex(double dagger) = 50.7 +/- 0.6 kJ mol(-1), Delta S-ex(double dagger) = -81 +/- 2 J K-1 mol(-1), and Delta V-ex(double dagger) = -10.4 +/- 0.5 cm(3) mol(-1). For the couple Co(diamsarH(2))(5+/4+) at 25 degrees C, the corresponding parameters are E degrees’ = +0.03(2) V and Delta V = +19.5 +/- 0.8 cm(3) mol(-1) (I = 0.20 mol L(-1)) vs Ag/AgCl/4 mol L(-1) KCl, k(ex) = 0.0149 +/- 0.0005 L mol(-1) s(-1), and Delta V-ex(double dagger) = -9.6 +/- 0.8 cm(3) mol(-1) (I = 0.39 mol L(-1)). The rate constants and volumes of activation are as expected for these electron transfer reactions on the basis of an adaptation of Marcus-Hush theory. Comparison with data for other cation-cation self-exchange reactions shows that low-spin/high-spin Co-III/II cage complexes behave normally in the Marcus-Hush context, despite the large change in spin multiplicity accompanying electron transfer, whereas nonencapsulated Co-III/II chelate couples deviate strongly from the norm, especially in Delta V-ex(double dagger) values which are anomalously strongly negative. This is attributed to structural changes associated with a Co-II spin equilibrium preceding electron transfer in the nonencapsulated chelates-changes which are suppressed in the cage environment. The alternative view, that the anomalous behavior of the Co chelates stems from nonadiabaticity, now seems untenable. Thus, spin multiplicity change per se does not inhibit electron transfer, i.e., such reactions are not "spin forbidden", but distortions that may accompany it in electron transfer between flexible complexes may retard the reaction and make a substantial negative contribution to Delta V-ex(double dagger).