Innovation is a complex trade-off between routinisation and change, between reliability and accountability of firms and timely adaptation. This innovator's dilemma confronts innovation theory with the question, how to align routinisation with innovation induced organisational change and consistent performance. Obviously it is a complex issue. Dominant innovation theory neglects this subject due to its pro-innovation bias, while evolutionary organisation and innovation theory give opposite perspectives on this problem. The adaptation perspective considers pro-active behaviour as the best condition for innovative performance, whereas the selection perspective advances inert firm behaviour as the best alternative to achieve successful innovations. Our research question focuses on the explanatory value of either the selection or the adaptation perspective for the innovative performance of industrial firms. Our empirical findings confirm the adaptation perspective and reject the selection perspective. Comparatively, firm behaviour involving the highest risks and uncertainties-e.g., high environmental dynamics and high levels of adaptive activity-contributes most significant to the explanation of innovative performance. Inert, risk averse behaviour, conversely, does not improve or even impedes innovative performance compared with other types of firm behaviour. (C) 2000 Elsevier Science B.V. All rights reserved.