A security policy database identifies the intended security policies within a network, a traffic generator provides test traffic that is configured to test each defined security policy, and a simulator simulates the propagation of this traffic on a model of the network. The model of the network includes the configuration data associated with each device, and thus, if devices are properly configured to enforce the intended security policies, the success / failure of the simulated test traffic will conform to the intended permit / deny policy of each connection. Differences between the simulated message propagation and the intended security policies are reported to the user, and diagnostic tools are provided to facilitate identification of the device configuration data that accounts for the observed difference. Additionally, if a network's current security policy is unknown, test traffic is generated to reveal the actual policy in effect, to construct a baseline intended security policy.