This paper discusses the approch to measuring historical thinking competencies developed in the projekt 'HiTCH: Historical Thinking - Competencies in History' (2012-15). Based on the so-calles 'FUER-Modell' of historical competencies, several tasks were developed and tested (N=2853). Besides presenting the key project results, this paper addresses three aspects: firstly, the extent to which such a quantitive test does not merely measure specifiv reading and/or general cognitive abilities, secondly, the distribution of correct solutions; and thirdly, the possible causes of different responce-frequencies, which are discussed on the basis of two sample tasks.