Energy efficiency (EE) has become an important design goal for future cellular systems. In this paper, we study the EE problem for limited-feedback coordinated beamforming systems providing real-time services with constant date rate. To maximize the EE, we jointly optimize the transmit power among users and the feedback bits of each user to quantize the desired and interfering channel directions, subject to the constraints of outage probability and number of feedback bits. An iterative optimization algorithm is proposed for alternating power and bit allocation by employing Yates’ framework. To further reduce the computational complexity, a suboptimal algorithm is proposed in closed form by decoupling the problem into subproblems under asymptotical analysis. Simulation results show that the suboptimal algorithm performs closely to the iterative optimization algorithm under high-quantization resolution. Moreover, the proposed two algorithms provide substantial EE gain over existing schemes with equal power and bit allocation, and the EE depends on the target data rate and outage probability of each user as well as the circuit power consumption at the base stations.