This paper deals with automatic generation control (AGC) of two unequal interconnected thermal areas considering the reheat turbines for thermal areas and appropriate Generation Rate Constraint (GRC). The response with GRC is compared with the analysis done without the Generation Rate Constraint. Although the frequency deviation is less with suitable controllers when the GRC is not considered, it is not the actual frequency deviation. When GRC is considered the actual frequency deviation can be found and then accordingly the controller is tuned. Particle swarm optimization (PSO) technique is used to simultaneously optimize the integral gains (KI), speed regulation parameter (Ri) and frequency bias (Bi) parameter. Most of the literature for AGC used classical approach based on integral squared error (ISE) technique, etc. for optimal selection of controller parameters. This is a trial and error method; extremely time consuming when several parameters have to be optimized simultaneously. The computational intelligence based technique like PSO is a more efficient and fast technique for optimization of different gains in load frequency control. MATLAB/SIMULINK is used as a simulation tool.