Research summary: Since Nickerson and Zenger (2002) proposed how vacillation may lead to organizational ambidexterity, large‐sample empirical tests of their theory have been missing. In this paper, we empirically examine the performance implications of vacillation. Building upon vacillation theory, we predict that the frequency and scale of vacillation will have inverted U‐shaped relationships with firm performance. We test our hypotheses using patent‐based measures of exploration and exploitation in the context of technological innovation and knowledge search. Managerial summary: Firms often shift their focus on technological innovation and knowledge search from seeking new and novel knowledge (i.e., exploration) to extending and refining existing knowledge (i.e., exploitation) or vice versa. We examine how the frequency and scale of firms vacillating between exploration and exploitation may affect their performance. We find that both too infrequent or too frequent changes and a too small or too large scale of changes are not desirable. Copyright © 2016 John Wiley & Sons, Ltd.