Subjective data on satisfaction with household income can be used to estimate equivalence scales and the cost of children. This avoids the well-knowm identification problems associated with estimating equivalence scales from consumption data. Using data from the German Socio-Economic Panel drawn in 1998, this paper compares several models that explain satisfaction with household income, family composition, and other background variables. Parametric as well as semi-parametric models are estimated and tested against more general alternatives, and equivalence scales based upon these models are compared.