In this paper a new class of cyber attacks against state estimation in the electric power grid is considered. This class of attacks is named false data injection attacks. We show that with the knowledge of the system configuration an attacker could successfully inject false data into certain state variable while bypassing existing techniques for bad data detection. In the preliminary section we consider the feasibility of such an attack and the necessary condition to successfully avoid detection. After that we show that with the knowledge of the system configuration, certain line flow measurements could be manipulated to lead to profitable misconduct. By controlling Regional Transmission Organizations (RTOs) view of system power flow and congestion, an attacker could manipulate the LMPs of targeted buses according to prior biddings. Also, in this paper we show the implementation of the false data injection attacks. The numerical example considered was applied to a malicious data detection algorithm that was designed on a microcontroller. The results demonstrated the effectiveness of injecting false data measurements into the state estimation of electric power grids.