In the present work, the performance of six convective parameterization schemes (CPSs) of the Weather Research and Forecasting model is tested for the intensity simulation of the cyclonic storms Mala, Sidr, Nargis, Aila, Laila, Jal and Thane which formed over the Bay of Bengal, with a special focus on Jal (formed in November 2010). It is found that the Grell-3, new Kain–Fritsch and old simplified Arakawa–Schubert CPSs provide higher intensity and give higher divergence in the upper-level convective precipitation rate, horizontal eddy viscosity, eddy viscosity of heat and total advective moisture tendency than other tested CPSs. It is noted that no single CPS can be assumed “best” for simulating all characteristics of a storm including intensity, track, rainfall amount, the spatial distribution of rainfall, and landfall (time and location). The study also incorporated some of the critically simulated parameters such as vorticity (shear, curvature and total), latent heating, relative humidity, equivalent potential temperature, cloud water content and cloud fraction in the intensity forecast and their importance in the cyclone forecast is discussed. The CPSs influence the forecasted convective precipitation rate of the storm and have a good correlation with the maximum surface wind, cloud fraction and upper-level divergence. © 2019 Royal Meteorological Society