Base hydrolysis of optically pure mer-exo(H)- and mer-endo(H)-[CoCl(dien)(dapo)]2+ (A and B (X = Cl), resp.: dien = N-(2-aminoethyl)ethane-1,2-diamine; dapo = 1,3-diaminopropan-2-ol, k(OH) = (1.134-0.09).10(5) M-1s-1 (A (X = Cl)), ko, = (1.18 +/- 0.11) . 10(5)M-1s-1 (B (X = Cl)); I = 1.0m (NaClO4 or NaN3), T = 298 K) is accompanied by retention of the mer-geometry and by full racemization (99 +/- 1%). It is shown that this is not due to racemization of either reactants or products. This result, together with the fact that both A and B yield the same mer-exo(H)/mer-endo(H)-product distribution, indicates the intermediacy of a pentacoordinate species II which is symmetrical (at least in the time average), viz. trigonal bipyramidal with a deprotonated ('flat') secondary-amine moiety. The H-exchange rates of the coordinated amine groups are consistent with this interpretation and indicate that loss of Cl- is the rate-determining step, in agreement with an S(N)1CB mechanism. The reactivity of the unsym-fac-exo(OH)- and unsym-fac-endo (OH)-isomers C and D, respectively, is in sharp contrast: base hydrolysis is 3 orders of magnitude slower, and the reaction is accompanied by some change of coordination geometry (C, 23%; D, 10%), some inversion of configuration (C, 15%; D, 19%); and much lower acceleration of hydrolysis in base (10(6) vs. 10(10)). Azide competition during base hydrolysis of the mer-isomers A and B is quite large (R = [CoN3]infinity/[CoOH]infinity[N3-] = 1.4 +/- 0.2m-1, I = 1.0m, T = 298 K) and indicates that the coordinatively unsaturated intermediate II is highly selective. The ratios of exo (H)- and endo (H)-azide competition products A and B (X = N3), respectively, immediately after the substitution reaction (kinetic control) are independent of the engaged epimer A or B: 31.7 +/- 0.9 % of B (X = N3) and 68.3 +/- 0.9% of A (X = N3, determined after ca. 10 . t1/2 of the base hydrolysis). This is in agreement with the effective site of deprotonation at the secondary(central)-amine group of dien, cis to the leaving group X, and with a common set of intermediates. Epimerization of A and B (X = Cl, N3) is shown to proceed solely via the pentacoordinate (base hydrolysis) intermediate II, viz. the direct route involving a six-coordinate deprotonated intermediate is immeasurably slow. For the hydroxo products A and B (X = OH), the direct route may compete with the H2O-substitution (exchange) path which can occur by an internal conjugate-base process. The kinetically controlled distribution of complexes A/B (X = N3) is different from the quasi-thermo-dynamic one (19.1 +/- 0.8% of B (X = N3) and 80.9 +/- 0.8% of A (X = N3)). This is consistent with the differences in the base-hydrolysis rates of the reactants (k(OH) (A (X = N3)) = (1.59 +/- 0.04) . 10(2)M-1S-1; k(OH) (B (X = N3)) = (2.89 +/- 0.22) - 10(2)M-1S-1). Various aspects of the investigated reactions are discussed on the basis of the widely studied reaction of base hydrolysis of pentaaminecobalt(III) complexes. Also, the structure and reactivity of the pentacoordinate intermediate II are discussed in relation to various current models.