The potential energy surface comprising ionized pyrimidine, 1·+, and eight of its hydrogen-shift isomers, as well as that of the corresponding neutrals was explored at a level of theory (B3LYP/TZVP) that has proven adequate for related species. The computations predicted that among the isomers there are four C4H4N2·+ distonic radical cations, 2·+– 5·+, of comparable stability to 1·+, and transition state calculations indicated that high barriers separate these stable ions. Thus, the ions 2·+– 5·+ should also be viable chemical species, and indeed mass spectrometry based experiments lead to the generation and characterization of three of the four, that is 2·+, 3·+, and 4·+, as stable ions in the gas phase. Ions 2·+– 4·+ were identified on the basis of their collision-induced dissociation characteristics in the mass spectrometer. The ions 2·+ and 3·+ obtained by dissociative electron impact ionization were subjected to neutralization–reionization mass spectrometry (NRMS). From collision-induced dissociation spectra of the intense NRMS survivor ions, it follows that the neutral ylide/carbene counterparts, i.e. pyrimidine-4-ylidene, 2, and pyrimidine-2-ylidene, 3, have lifetimes of at least microseconds in the rarefied gas phase. The interpretation of the experimental observations that 2 and 3 are viable chemical species in gaseous environs was supported computationally. According to the calculations the neutral isomers 2 – 5 each represent a minimum separated by high hydrogen-shift barriers, although situated some 50 kcal/mol higher in energy than 1, pyrimidine itself. However, molecules 4 and 5 remained elusive since ions 4· +, only obtainable by a collision-induced dissociation process, were not amenable to NR experiments and a viable strategy to produce a beam of pure ions 5·+ could not be realized.