A recent innovation in empirical academic research is the use of online labour markets as a source of data. One such market, Amazon's Mechanical Turk ("MTurk") has been used by studies published in high-quality accounting journals to source participants. Given the traction of this data source in high-calibre publications, it is timely to assess its current impact and future potential for accounting research. This paper examines the extent of adoption of MTurk as a data collection tool in leading accounting journals and specifically considers its adoption and suitability for survey research. Findings reveal that the use of MTurk in high-quality accounting publications is gathering momentum, with approximately the same number of articles published/accepted in 2019 as the total number of articles published in the preceding seven years. However, it is also found that nearly all the journal articles reviewed adopted MTurk for experimental research with only a limited presence in survey research. The study contributes to the literature by providing a comprehensive review of the adoption of MTurk in high-quality accounting journals by frequency, research method, and research participant type. Further, it analyses the unique methodological concerns that MTurk poses for survey-based accounting research, thereby providing researchers with guidance on its potential future usefulness and pitfalls to avoid. The paper concludes that difficulties in the availability of, and screening for, specific groups of participants may limit its potential for survey research until online labour market platforms are developed further.http://bit.ly/mturk1519, accessed in June 2020.