A sub-optimal control strategy for space heating system is developed. The algorithm utilizes a reference signal and a dynamic gain factor to update the control sequence. The reference signal was computed using steady-state optimal tank temperature based on the forecasted daffy minimum temperature. The controller tracks this temperature and thereby minimizes the energy ~nput to the heating system. The dynamic gain factor was introduced to account for the fluctuations in room air temperature and the heating system dynamics. Both these parameters have the net effect of reducing energy consumption and maintaining desired space temperatures. The operation of the controller is simulated and its response characteristics are analyzed. Also, the controller performance in terms of annual energy consumption is compared with a conventional on-off control system.