The structures of carbocations formed in the ionization of 3-methyl-2-butyl precursors were investigated by high level ab initio MO calculations. The relative stability of the 3-methyl-2-butyl cation (3) increased upon placement in a dielectric medium using SCRF calculations, but the 2,3-dimethyl-1-protonated cyclopropanes ( 5) were still more stable, as they were in vacuum. Introduction of negative point charges in single-point calculations on the structures of isolated ions made the open ion more stable than the bridged isomers and increased the stability difference favoring trans-5 over cis-5. The structure and stabilities of 3 and trans-5 in ion pairs were examined with two counterions successfully used in previous studies, trihydrofluoroborate (FBH 3 -, A) at short interionic distances (d) and dihydrolithiate (LiH 2 -, B) at longer distances. Optimization (MP2/6-31G*) of trans-5.A with the anion free to move in a plane parallel to the C1C2C3 plane at d ) 2.3 Å gives 3.A as the only energy minimum. The position of the anion is determined by electrostatic interaction with C2 and hydrogen atoms at C1 and C5. At 2.6 Å, trans-5.A also opens, but the anion in the resulting 3.A moves toward C1 and forms 3-methyl-1-butene by elimination. Rotation of the cation to the conformation in which the hydrogen at C3 of 3.A faces the anion at d ) 2.6 Å leads to 2-methyl-2-butene. Optimization at longer interionic distances with the anion (B) placed above C2, along a line perpendicular to the C1C2C3, plane, shows 3.B as the only stable entity up to d ) 3.25 Å. Both 3.B and trans-5.B were optimized at d ) 3.3 Å, where 3.B was more stable by 2.11 kcal/mol (MP4SDTQ(FC)/6-31G**//MP2(FC)/ 6-31G** + ZPE). At greater separation, rotation of C2-C3 brings one hydrogen at C5 closer to the anion. At d ) 3.8 Å, ring closure in 3.B occurs with participation of C5, rather than C4 which should bridge in an anchimerically assisted ionization. Another orientation of the anion was tested, placing B on a line perpendicular to the C2,C3,C5 plane. Optimization of trans-5.B at d ) 4.0 Å led to 3.B, but trans-5.B could be optimized at d ) 4.5 Å, where it was 7.1 kcal/mol less stable than 3.B. The calculations indicate that there should be no anchimeric assistance by either methyl or hydrogen upon ionization of a 3-methyl-2-butyl precursor to ion pairs, in agreement with the previous study of this process in trifluoroacetic acid.