A number of artificial and natural systems can be modeled as hybrid models in which continuous and discrete variables interact. Such hybrid models are usually challenging to analyze and control due to the computational complexity associated with existing methods. In this paper, the novel modeling formalism of Guarded Flexible Nets (GFNs) is proposed for the modeling, analysis and control of hybrid system. A GFN consists of an event net that determines how the state changes as processes execute, and an intensity net that determines the speeds of the processes. In a GFN, the continuous state is given by the value of its state variables, and the discrete state is given by the region within which such variables lie. GFNs are shown to possess a high modeling power while offering appealing analysis and control possibilities.