A cloud-resolving model is used to study the effects of a vertical temperature dipole on convective cloud development. Such dipole anomalies, with a warm-above-cool structure in the troposphere, are known to be forced by mesoscale convective systems (MCSs) in the Tropics. The experiments involve letting convection develop in perturbed initial soundings with open lateral boundary conditions. Convection is driven solely by surface fluxes. In the control run, a field of deep convection ensues. With a strong dipole anomaly that is warm in the upper troposphere, no clouds ascend beyond the middle troposphere. In this case, cumulus congestus clouds strongly moisten the midtroposphere with relative humidity increases by up to 24% by the end of the 6‐h simulation. With a half-strength anomaly, a mixed population results: mainly middle-topped congestus clouds, but with some intermittent deep cells. The partitioning between cloud types is somewhat sensitive to model resolution, with a change from 1- to 0.5-km grid spacing resulting in relatively more congestus clouds and fewer deep cells.