Research summary: We use a sample of Singapore firms to replicate Barkema and Vermeulen's (1998) study of international expansion through start‐ups or acquisitions by Dutch firms. We discover that the authors misinterpreted the regression coefficients for hypothesis testing and only two of their four hypotheses were actually tested. For these two hypotheses, one is not supported in either their study or ours, while the other is supported in their study but not ours. For the remaining two hypotheses we find support for one of them, which is concerned with the curvilinear effect of product diversity on the mode of expansion. In sum, the original study claims that all four hypotheses are supported, whereas only one is supported in the replication. More specifically, the former results, including the effects of the independent and control variables, are largely not generalizable to the latter. Managerial summary: Barkema and Vermeulen's (1998) study investigates the international expansion by Dutch firms during the period from 1966 to 1994. Their results indicate that whether a firm expands through setting up a greenfield operation or acquiring an existing operation is affected by the diversity of the firm's product lines, the diversity of the countries to which it has expanded, and how far the expansion is related to its existing business. We replicate their study using a sample of Singapore firms for the period from 1980 to 2000. Our results show only an effect of a firm's product diversity on its mode of international expansion. Our study clearly indicates the risk of drawing managerial implications from the results of a single study. More replication studies are needed for establishing a solid theoretical foundation to inform management decisions. Copyright © 2016 John Wiley & Sons, Ltd.