In this study, based on simulations of a two-dimensional multicell storm under a ground-layer upshear (Uz< 0) by a mesoscale numerical model, a new mechanism of cell regeneration and development within the multicell storm at the "less than optimal shear" state.is proposed.In the presence of a ground-layer upshear, the circulation associated with the surface cold pool is not counteracted by that associated with the ambient wind shear, and the density current extends out faster, making the multicell storm stay at the "less than optimal shear" state. As a result, a new cell is triggered by the strong vertical perturbation ahead of the mature convection, rather than by the split-up from the updraft at the leading edge of the surface cold pool as well as the gust front. The latter is the mechanism at the "optimal" state proposed by Lin et al. in 1998. In the new mechanism, the regenerated cell grows fast with the incident warm moist air from the upstream of the multicell storm, and tends to cut off the moist airflow into the mature convection at its western sector. Consequently, the mature convection would weaken, be replaced, and eventually decay.Actually, these two different mechanisms come into play in a way depending on the relationship between the circulation of the low-level shear and that of the cold pool. When the circulation of the cold pool is stronger than that of the wind shear, the multicell storm is at the "less than optimal shear" state, and the new convective cell is produced by the disturbance ahead of the mature cell. When the circulation of the cold pool is weaker, the cell regeneration is dominated by the mechanism at the "optimal" state, and the new cell is split from the gust front updraft. Therefore, these two mechanisms are not contradictive. With a moderate ground-layer upsheax, they can alternately operate within a multicell storm.