The present paper employs spatial econometrics techniques to discriminate empirically between various economically plausible spatial patterns of interregional knowledge spillovers between west German planning regions in the 1990s. In general, interregional spillovers are found to contribute significantly to regional knowledge production. Due to fairly high spatial transaction costs, however, only a small fraction of the knowledge available in neighboring regions actually spills over. Consequently, the absolute contribution of ‘foreign’ knowledge to a region's innovative performance is quite low. Moreover, only regions with low R&D density benefit from interregional spillovers. For regions with high R&D density they seem to be negligible. One reason for this may be some sort of self-sufficiency in the R&D centers where researchers may have fewer incentives to consult researchers in other regions. Another reason for this may be the dominance of unidirectional knowledge flows from technological leaders to followers.