Automating business processes has been a popular business strategy in the past few years. These processes were modeled using an imperative modeling language. A known issue of this way of modeling is that it is dificult to model exible parts of processes, since every sequence needs to be modeled explicitly. Flexible parts or parts that need user input of a process need to be made structural in order to be able to automate them. On the other hand, with a declarative modeling language it is not convenient to model structured parts of a process. A modeling approach that is able to model both the structured and exible parts does not exist. This thesis presents a hybrid process modeling approach that addresses this issue. The hybrid modeling language consists of a combination of the imperative modeling language BPMN and declarative modeling language Declare. It allows the user to jointly model exible and structured parts of a process. A method is developed to create a hybrid process from a process description and guidelines are created to help the user decide between modeling a group of activities using the imperative or declarative part of the hybrid modeling language. An empirical experiment is conducted to evaluate the set of guidelines. The analysis of the results indicates that these guidelines are found useful, but the `ease-of-use' aspect needs improvement.