This presentation offers some considerations arising from the empirical results of Stephen Pumfrey’s corpus-linguistic analysis of the use of ‘experimental’ and ‘experiment’ in early modern printed sources. Broadly, the research confirms Peter Harrison’s recent claim that these terms were current in religious and theological literature, before appearing in works concerned with the study of the natural world. But the picture is more complex than Harrison allows, and those complexities raise problems which he did not consider. In particular, Harrison does not notice that usage of these terms strongly suggests an association with radical sectarian views—those views decried as ‘enthusiastic’ by more mainstream thinkers. Consequently, there is a dramatic falling off of talk of experimental religion in the 1660s at just the point when ‘experimental’ is taken up by natural philosophers. Immediately, this raises the question as to how experimental philosophers avoided charges that their experimental philosophies must be tainted with the same kind of ‘enthusiasm’ as seen in the experimental religionists. An obvious response to this is for the historian to suggest that previous, and on-going, scholarship on the origins of the experimental method (from Roger Bacon to Francis Bacon) shows that the term ‘experimental’ must have had a separate and well-understood significance which enabled readers in the 1660s to recognise that natural philosophers were using the term in this sense, rather than in the potentially subversive sense of the religious ‘enthusiasts’. But, if the terms ‘experiment’ and ‘experimental’ had this separate, parallel, usage in natural philosophy and related traditions (natural magic, mathematics, and medicine), why do they not show up in corpus-linguistic surveys? Furthermore, we can turn this around and ask what are the implications of the lack of usage of ‘experiment’ and ‘experimental’ in natural philosophy before the 1660s for scholarship in the history of science, which has traced the origins of the experimental method back to Francis Bacon and even earlier? The presentation will address these and other problems arising from this corpus-linguistic research, and may even offer some tentative answers.