Dose assessments based on multimedia exposure models require intermedia-transfer factors (ITFs) as input. These factors relate contaminant concentrations in an environmental medium (the source medium such as air. water, or soil for which measurements are available) to the exposure medium (personal air, tap water, and food products) with which humans have contact. In this paper, I use QSAR as applied to partition coefficients between soil and plants, air and plants, and between animal diet and lipid tissues as case studies for characterizing the precision of ITF estimations. These partition coefficients form the basis of total dose models required for realistic exposure/risk assessments. The quantitative relationship between ITFs and chemical properties are developed along with the estimation errors associated with these relationships. Many of these correlations have large estimation errors that limit their reliability as applied in risk assessment. I examine factors that contribute to these estimation errors and consider a “value of information” approach for increasing the precision of these estimation methods.