The present study investigated the relationship between the use of different internet applications and research productivity, controlling for other influences on the latter. The control variables included dummies for country, discipline, gender and type of organization of the respondent; as well as variables for age, recognition, the degree of society-related and career-related motivation for research, and the size of the collaboration network. Simple variance analyses and more complex negative binomial hurdle models point to a positive relationship between internet use (for personal communication, information retrieval and information dissemination) and research productivity. However, the results should be interpreted with caution as it was not possible to test the role of the internet against other pre-internet tools which fulfil the same functions. Thus instance it may not be the use of e-mail per se, but the degree of communicating with colleagues that makes a productive scientist.