The present work aims to illustrate the criteria used for the determination of probabilities of selection of first and second stage units, given the available information, within the TLS design implemented for the survey of the Relevant National Interest Project (PRIN) on mobility of regional incoming tourism in Sicily and Sardinia. First and second stage units selection probabilities are necessary for the implementation of Hansen-Hurwitz class estimators for the estimation of the average number of stops made by tourists in Sicily, for the estimation of the share of un-observed tourism, and for the sampling errors associated.