Abstract: Memristor plays an important role in modeling nonlinear circuits and systems. Based on the proposed smooth quadratic generic memristor, this paper proposes a memristor-based Lorenz chaotic system. Different from the chaotic system on account of memristor feedback, this system takes a variable of the original Lorenz chaotic system as an inner state variable of memristor, so as to ensure that the system dimension does not increase. Stability analysis shows that the system has the same equilibrium point and stability as the original Lorenz system, namely, one unstable saddle point and two unstable saddle foci. By means of bifurcation diagram, Lyapunov exponent spectra, and phase plot, the dynamics of the proposed memristive system are revealed. The simulated results show that the memristive Lorenz chaotic system possesses coexisting bistable mode and self-similar bifurcation structures. What is more interesting is that the amplitude of the system can be regulated by changing the inner parameters of the generic memristor. Finally, the equivalent circuits of memristor and memristive system are designed and also synthesized by analog components. Simulation results confirm the correctness of the numerical simulations.