As a form of non-structural data, textual big data has been deeply influencing accounting and finance research in recent decade. This influence manifests in two strands of literature: The first centers on information, such as applying textual analysis in measuring the quality (such as readability) and quantity of information (the information content of financial document), information disclosure and anomalies among other related issues. The second strand is mainly on applying textual analysis to construct new measure, such as competition, innovation and Economic Policy Uncertainty and so on. This paper first surveys the two strands of literature and then analyzes the advantages and disadvantages of textual analysis. Lastly we point out the opportunities and challenges in applying textual analysis in accounting and finance research.