NeuroD/BETA2 (referred to as NeuroD hereafter) is a basic helix-loop-helix (bHLH) transcription factor that is required for the development and survival of a subset of neurons and pancreatic endocrine cells in mice. Gain-of-function analyses demonstrated that NeuroD can (i) convert epidermal fate into neuronal fate when overexpressed in Xenopus embryos, and (ii) activate the insulin promoter in pancreatic beta cell lines in response to glucose stimulation. In glucose-stimulated INS-1 pancreatic beta cells, mutations of S259, S266, and S274 to alanines inhibited the ability of NeuroD to activate the insulin promoter. Phosphorylation of those serine residues by ERK1/2 was required for NeuroD activity in that assay. To determine whether the same residues are implicated in the neurogenic activity of NeuroD, we mutated the conserved S259, S266, and S274 of Xenopus NeuroD to alanines (S259A, S266A, and S274A), and performed an ectopic neurogenesis assay in Xenopus embryos. In contrast to what has been observed in the pancreatic beta cell line, the S266A and S274A mutant forms of Xenopus NeuroD displayed significantly increased abilities to form ectopic neurons, while S259A had little effect. In addition, S266A and S274A of Xenopus NeuroD resulted in increased accumulation of protein in the injected embryos while the corresponding mutations on mouse NeuroD did not have the same effect in an insulinoma cell line. Our results demonstrate that the consequence of NeuroD protein modification is context-dependent at both the molecular and functional levels.