This paper develops a mathematical model for cognitive performance of a tactical air traffic controller in an en-route ATC context. The aim of this model-based approach is to enable the evaluation of both accident risk and aspects like cognitive workload and effectiveness of ATC in managing air traffic situations safely.Use is made of human error modelling, HoUnagel's cognitive mode model and Wickens Muhiple Resources model.The paper describes how these psychological sub-models are combined into a single model of ATCo cognitive performance, and how the interaction of these human sub-models with the technical sub-systems is brought into account. The model is applied to an exemplary ATM scenario which consists of two parallel lanes of opposite irection traffic at the same flight level,and where the air traffic controller has no automation support tools like Short Term Conflict Alert (STCA) or Flight Path Monitoring.The obtained results for this conventional ATC situation are compared to those previously obtained under the hypothetical assumption that a tactical air traffic controller reacts to aircraft deviations in case of an STCA alert only.