Résumé : Goma city, at the eastern border of DRCongo, is highly exposed to natural hazards, especially from Nyiragongo volcano, located directly North of it. In January 2002, the city centre of Goma was devastated by lava flows and several thousands of people were temporarily displaced. Defining and quantifying population vulnerability to natural hazards, and lava flow hazards in particular, is a crucial element to evaluate and manage the risk. This paper aims at assessing the vulnerability of the population facing volcanic hazards in Goma, and its spatial variation across the city, in order to support volcanic risk prevention and management at the local levels. In this data scarcity context, two parallel methodologies are tested based on data collected through a large-scale household survey: the Social Vulnerability Index (SoVI) as defined by Cutter et al. (2003) based on a statistical data reduction; the other using fewer significant indicators in order to develop an Operational Vulnerability Index (OVI). Results show that the spatial distribution of the vulnerability levels with both approaches is quite similar, but the construction of an OVI can help to communicate the message more easily to political authorities for risk management actions – e.g., to target neighborhoods where to develop priority prevention programs – but also in terms of spatial urban planning – e.g., to identify areas where to act. Population vulnerability assessment, together with the lava flow hazard invasion probability and population exposure, is one of the crucial steps towards the lava flow risk assessment.