Motivation: Microarray technology provides cell-scale expression data; however, analyzing this data is notoriously difficult. It is becoming clear that system-oriented methods are needed in order to best interpret this data. Combining microarray expression data with previously built pathway models may provide useful insight about the cellular machinery and reveal mechanisms that govern diseases. Given a qualitative state - transition model of the cellular network and an expression profile of RNA molecules, we would like to infer possible differential activity of the other molecules such as proteins on this network. Results: In this paper an efficient algorithm using a new approach is proposed to attack this problem. Using the regulation relations on the network, we determine possible scenarios that might lead to the expression profile, and qualitatively infer the activity differences of the molecules between test and control samples. Availability: This new analysis method has been implemented as part of a microarray data analysis component within PATIKA (Pathway Analysis Tool for Integration and Knowledge Acquisition), which is a software environment for pathway storage, integration and analysis. Facilities for easy analysis and visualization of the results is also provided. Contact: http://www.patika.org.