It is proven that the observed return rates on capital have an upward bias if firms are producing with unobserved intangible capital. Using a comprehensive firm level database for Germany, this theoretical preposition is supported empirically. Furthermore, by making unobserved intangible capital observable, dispersion in return rates is dramatically reduced. The results support the assumption that a considerable part of the observed dispersion in return rates among firms is attributable to unobserved capital formation in intangible capital.