The deformation behaviour of a hot rolled micro-alloyed steel bar of grade 38MnVS6 was examined using an FEM model during the initial passes in a blooming mill, as a function of three different pass schedules, roll groove depth, collar taper angle and corner radius. The simulations predicted the effective strain penetration, load, torque, fish tail billet end shapes, and metal flow behaviour at a chosen temperature, mill rpm and draft. The model predictions were validated for typical groove geometry and a typical pass schedule. Lower collar taper angle, lower corner radius and higher depth of groove in hot rolling enabled achievement of higher strain penetration, higher mill load and lower fish tail formation. The present study establishes the capability of the model to improve the internal quality of the rolled billet as measured by effective strain which was corroborated to the rolled bar macrostructure and microstructure. The model enables yield improvement by the choice of draft to minimise fish tail losses. The surface quality is improved by the ability to avoid fin formation that occurs at certain conditions of rolling. Thus, the groove geometry, roll pass schedule and rolling mill parameters and temperature can be optimised for best product quality and yield.