It is argued that the observed return rates on capital at firm-level have an upward bias if firms are producing with unobserved intangible capital. Using EUKLEED, a comprehensive firm level data base for Germany, this theoretical preposition is proved empirically. Furthermore, making unobserved capital observable the dispersion in return rates reduces dramatically. The results clearly support the assumption that a considerable part of the observed dispersion in return rates among firms can be contributed to unobserved capital formation in intangible capital. Firms with high input in intangibles also have an above average observed rate of return. However, the question to what extent a more intense use of intangibles can be the cause for higher return rates in the sense of both the monopoly-based and the innovation-based explanations is not answered.