Although tourism expenditure has long been a pertinent topic in studies dealing with cultural tourism, its importance in recent years has become even more marked due to the consequences of low-cost tourism that many destinations are suffering. This need has been further aggravated by the impact of the ongoing Coronavirus pandemic. Now, the only sustainable means of increasing destination revenue is to generate greater financial profitability of tourism-related activities, as opposed to increasing tourist volumes. The present study analyzes the determinants of the likelihood of tourist spending in an urban-cultural destination whose economic sustainability is being threatened by low-cost tourism. To this end, all the tourism activities of the sample (672 in total) were recorded in real time during the stay via a purpose-designed mobile application. Given the nested structure of the sample, multilevel modeling was used: the characteristics of the different tourism activities were used as predictor variables, while characteristics of the tourist were used as control variables. Regarding the former, the results suggest that spending is more likely at the beginning of the stay and in relation to non-cultural tourism activities (restaurants, shopping, transport, etc.). An interaction effect between activity location and timing (beginning vs. end of stay) was also demonstrated: at the beginning of the stay, the greater likelihood of spending was related to services or attractions outside the city center; and, toward the end, spending patterns become more static, based close to the city center. The aim is to explain the probability of tourist expenditure at each spending opportunity, thus contributing to the current knowledge of total tourist spending. Knowledge of tourist expenditure patterns is a prerequisite for raising profitability-per-tourist when increasing visitor volumes is not an option.