The effectiveness of any, methodology used to identify, hazards in chemical processes affects both safety, and economics. To achieve maximum safety at minimum cost, it conservative, but realistic, analysis must be carried out. An approach to hazard identification is proposed based on a detailed process model which includes nonlinear dynamics and uncertainty. A new modeling framework, the region-transition model (RTM), is developed, which enables the simulation of regions of the operating space through art extension of the hybrid state transition system formalism. The RTM is illustrated oil a nonlinear batch reactor with parameter uncertainty. A safety-verification algorithm identifies regions of the input space (initial conditions and external inputs) which guarantee safe operation. The algorithm is successfully applied to three examples; a tank with overflow and underflow, a batch reactor with an exothermic reaction, and a CSTR with feed preheating.