Lately two modules which aim to improve the modeling of human reasoning have been introduced into the ACT-R cognitive architecture – the Human Reasoning Module (HRM) and the Pre-Attentive and Attentive Vision Module (PAAV). This is a first attempt to create a domain-specific infrastructure as an extension of a cognitive architecture. The HRM defines the basic functionality of two different theories in human reasoning – the mental model theory and the mental logic theory. Thus, it is a step towards a “unified theory of human reasoning”. In this article we use a model of the continuity effect in spatial reasoning to evaluate this approach. The results show that the HRM is a clearly more convenient way to define ACT-R models for reasoning domains. However, as a unified theory of reasoning improvements are necessary before quantitative measures of all aspects in a reasoning task can be predicted.