Recently, research has focused on attitudes towards inclusive education, and the majority of studies use questionnaires to measure this vital variable. In two consecutive experiments, we showed that attitudes towards inclusive education are not stable but instead are significantly influenced by social context. We manipulated information on the organisation conducting a survey regarding attitudes of participants towards inclusive education. The results show that the attitude of the organisation conducting the survey – as perceived by the participant – outperforms well-documented variables (e.g. sex, age, and contact to a person with disability) in predicting the attitudes of the participant towards inclusion. This one variable explains as much variance as all other predictors combined. We argue that social desirability is a neglected issue in research on attitudes towards inclusive education. Our findings challenge the validity of numerous studies on this topic because they show a positive bias in the attitudes of participants when they were surveyed by a university. Thus, we outline the first steps to reduce social desirability-induced validity problems in research on attitudes towards inclusion.