It is known that nonlinear sampled-data control sometimes shows chaotic behavior. In this paper, chaos in an adaptive model predictive control system has been studied. The closed-loop behavior of the system is represented by a discrete dynamical system with a rational function of state variables. Several characteristics of the chaos in the system have been identified by providing bifurcation diagrams, attractors in the phase space, stable and unstable manifolds, and Lyapunov exponent.