The purpose of this article is to reflect on how different methods for building future scenarios are selected, according to the nature of the scenarios (quantitative, qualitative or mixed) and the characteristics of the exercise being conducted (regional coverage, time horizon, number of scenarios, and sponsor). In total, the authors reviewed 2,603 abstracts of articles published in indexed journals over the period of 2003-2013, 1,009 of which were published in Futures, 1,169 in Technological Forecasting and Social Change, and 425 in Foresight; with the goal of selecting those articles where exercises about future scenarios were presented, and evaluating the frequency of use of the methods based on their nature and the characteristics of the exercise. The various combinations proposed are presented graphically to respond to the purpose of the research, which finds that the nature of the method has a great influence and that the characteristics have little or very little impact on the choice of methods.