This article meets methodological and theoretical difficulties in using social indicators to assess vulnerability to climate change and contributes to enhance methodological strategies for addressing this phenomenon. Most of the studies of vulnerability to climate change define vulnerability as an expression of three dimensions: the exposures of the systems to climatic events, the sensitivity to these risk conditions and the adaptive capacity to cope with the impacts. These three dimensions enable the operationalization of the concept trough biophysical and social indicators whose main purpose is to quantify the phenomenon and establish causal relationships between events and resources. From a different wiew of this conceptualization, another way of understanding vulnerability is to understand it as a dynamic process strongly shaped by social, cultural, economic, political and institutional conditions. From this approach, climate change is suposed to impact on previous and long-term vulnerabilities mainly defined by access to resources. Vulnerability is then defined as a complex process which does not support quantifications or generalizations between causes and effects as those proposed in indicators systems. Based upon a research experience in Mendoza, Argentina, and Chinchiná, in Colombia this article demonstrates the limitations of the use of indicators of social vulnerability to applied social research. Through the analysis of public statistical data in Argentina and Colombia and in-depth interviews in the study areas, the authors reconstruct the difficulties faced in the building process of these indicators to demonstrate the weaknesses of each indicator to account the complexity of vulnerability. Income, land tenure, access to water, technology, infrastructure, technical assistance, education and associativism indicators particularly discussed. At the end, the authors presents an alternative proposal to adress the complexities of vulnerability. [ABSTRACT FROM AUTHOR]