Complex dynamics associated with multistability have been studied extensively in the past but mostly for low-dimensional nonlinear dynamical systems. A question of fundamental interest is whether multistability can arise in high-dimensional physical systems. Motivated by the ever increasing widespread use of nanoscale systems, we investigate a prototypical class of nanoelectromechanical systems: electrostatically driven Si nanowires, mathematically described by a set of driven, nonlinear partial differential equations. We develop a computationally efficient algorithm to solve the equations. Our finding is that multistability and complicated structures of basins of attraction are common types of dynamics, and the latter can be attributed to extensive transient chaos. Implications of these phenomena to device operations are discussed.