Abstract-Data flow analysis (DFA) technique is used toanalyze program into data variables and identify data flow operation on these variables. This basic information can be used to identify data dependencies, test data, program execution paths and hence it is helpful in the testing process to verify expected behavior of the system. In terms of executable UML, the models are based on action languages which internally consist of variables and specify data flow operation on these variables. Therefore, DFA of action languages is essential to analyze UML models in term of data flow and use the information to verify formal correctness (expected behavior) of the system. In our proposed approach, we are using action language for fUML (ALF) to design and analyze executable UML model. Analytical results show that DFA of executable models provides the precise execution flow of variables which are used to identify data dependencies and verification of system expected behavior from its abstract model.