Tourism statistics are one of the key sources of information for economists, public officials and tourism decision-makers. The aim of the present paper is to describe and critique the methodological difficulties encountered when approaching statistical studies in tourism. The case of hidden tourism in island communities is used to illustrate that in tourism statistics there exists a lack of clarity and convention concerning definitions, procedures, measurement and analytic approaches. The conclusions and the study implications should help tourism authorities and tourism statisticians to better define and standardise methodological and measurement approaches and practices and to more effectively direct research activities towards new approaches to difficult research questions.