This paper explores to what extent interpretations of individual primary carcinogenicity data differ between different risk assessors, and discusses possible reasons for such differences as well as their impact on the overall risk assessment conclusions. For this purpose 29 different TCE carcinogenicity risk assessments are used as examples. It is concluded that the TCE risk assessors surprisingly often interpret and evaluate primary data differently. Two particular reasons for differences in data interpretation are discussed: different assessments of statistics, and different assessments of whether the results obtained in bioassays have toxicological relevance. Differences in the interpretation and evaluation of epidemiological data are also explored and discussed.