This paper aims to offer insights into the usability, acceptance and limitations of e-readers with regard to the specific requirements of scholarly text work. To fit into the academic workflow non-linear reading, bookmarking, commenting, extracting text or the integration of non-textual elements must be supported. A group of social science students were questioned about their experiences with electronic publications for study purposes. This same group executed several text-related tasks with the digitized material presented to them in two different file formats on four different e-readers. Their performances were subsequently evaluated by means of frequency analyses in detail. Findings - e-Publications have made advances in the academic world; however e-readers do not yet fit seamlessly into the established chain of scholarly text-processing focusing on how readers use material during and after reading. Our tests revealed major deficiencies in these techniques. With a small number of participants (n=26) qualitative insights can be obtained, not representative results. Further testing with participants from various disciplines and of varying academic status is required to arrive at more broadly applicable results. Practical implications - Our test results help to optimize file conversion routines for scholarly texts. We evaluated our data on the basis of descriptive statistics and abstained from any statistical significance test. The usability test of e-readers in a scientific context aligns with both studies on the prevalence of e-books in the sciences and technical test reports of portable reading devices. Still, it takes a distinctive angle in focusing on the characteristics and procedures of textual work in the social sciences and measures the usability of e-readers and file-features against these standards.