Nowadays, almost everyone uses the World Wide Web (WWW) to search for information of any kind. In education, students frequently use the WWW for selecting information to accomplish assignments such as writing an essay or preparing a presentation. The evaluation of sources and information is an important sub‐skill in this process. But many students have not yet optimally developed this skill. On the basis of verbal reports, eye‐tracking data and navigation logs, this study investigated how novices in the domain of psychology evaluate Internet sources as compared to domain experts. In addition, two different verbal reporting techniques, namely thinking aloud and cued retrospective reporting, were compared in order to examine students' evaluation behaviour. Results revealed that domain expertise has an impact on individuals' evaluation behaviour during Web search, such that domain experts showed a more sophisticated use of evaluation criteria to judge the reliability of sources and information and selected more reliable information than domain novices. Furthermore, the different verbal reporting techniques did not lead to different conclusions on criteria use in relation to domain expertise, although in general more utterances concerning evaluation of sources and information were expressed during cued retrospective reporting. Lay Description What is already known about this topic: When searching the Web, students assess the reliability of information and sources in a spontaneous manner. Students tend to use superficial criteria when judging the trustworthiness of web‐based information and sources. Little is known about the differences between domain experts and novices in terms of the ways they spontaneously evaluate web‐based information and sources. What this paper adds: Domain knowledge influences evaluation behaviour, with greater domain knowledge leading to more reliable selection of sources. A comparison of techniques for measuring evaluation behaviour when searching the Web for information reveals that different techniques, such as thinking aloud and cued retrospective reporting, have different pros and cons whose implications depend on the research questions being addressed. Implications for practice and/or policy: In formal education, more attention should be paid to equipping students with the requisite skills in searching for and evaluating web‐based information. Embedded process‐oriented instructional designs are needed that foster the appropriate evaluation behaviours.