By comparing parallel field and survey experiments testing compliance with international standards on corporate transparency, we highlight potential problems in the external validity of survey experimental designs. We performed a field experiment using deception in which we requested an anonymous business incorporation from nearly 4,000 corporate service providers in more than 180 countries. Subsequently, we conducted a survey experiment with the same providers using similar treatment conditions, but with informed consent to participate in a research study. Comparing responses and response rates corroborates-from a new angle and with additional implications-survey researchers' caveats about selection bias and social desirability. Our conclusions on the relative external validity and different substantive results produced by different experimental designs constitutes an important cautionary note given the increased popularity of survey experiments within international relations and political science more generally.