Quantum-chemical calculations predict that synthetically accessible cyclic four-membered, four-pi-electron ylides could be used as building blocks for the realization of hitherto unknown N-heterocyclic boron(I) carbenoids. The boron(I) carbenoids proposed in this work possess the largest computed singlet-triplet separations known to date, which are comparable to those of the corresponding aluminum(I) analogue computed at the same level of theory. Furthermore, they owe their stability not only to the substantial transfer of electron density from nitrogen to boron atoms but also to the presence of thermodynamically robust ylidic bonds. On the basis of their computed proton affinity and carbonyl stretching frequencies, they may be considered as promising ligands for transition-metal complexes.