A signal transduction pathway (STP) is a cascade composed of a series of signal transferring steps, which often activate one or more transcription factors (TFs) to control the transcription of target genes. Understanding signaling pathways is important to our understanding of the molecular mechanisms of disease. Many condition-annotated pathways have been deposited in public databases. However, condition-annotated pathways are far from complete, considering the large number of possible conditions. Computational methods to assist in the identification of conditionally activated pathways are greatly needed. In this paper, we propose an efficient method to identify conditionally activated pathway segments starting from the identification of conditionally activated TFs, by incorporating protein-DNA binding data, gene expression data and protein interaction data. Applying our methods on several microarray datasets, we have discovered many significantly activated TFs and their corresponding pathway segments, which are supported by evidence in the literature.