Background In policy health impact assessment (HIA), quantifying and modelling accurately multivariate and complex realities is crucial to informed and evidence based decision- making processes. Multivariate statistical methodologies and data mining techniques represent a useful and powerful contribution. Our ongoing research aims to explore the potential added-value of multivariate statistics in each step of HIA. Analysing the impact of RSI (Social Welfare Payment for Inclusion) social policy allows us to evaluate its potential consequences on health and health care inequalities. Methods A case-control study was conducted in the Lisbon council (June 2007–March 2008), with a random sample of 1513 women of fertile age, divided into three groups: 499 women considered very poor, from the RSI beneficiaries of SCML (Santa Casa da Miserico ́ rdia de Lisboa)—Group 1; 1014 controls, including 507 poor women from other SCML beneficiaries—Group 2; and 507 non-poor women from four Health Centers in Lisbon Council—Group 3. Data were collected by personal interview (semi-quantitative question- naire). A total of 1054 women answered the questionnaire (61%, Group 1; 58%, Group 2; 90%, Group 3). Multivariate statistical data analysis included clustering models, multiple correspondence analysis and structural equation models. Results Applying multivariate statistical data analysis identified different profiles between women of the three groups. Preliminary results show distinct representations of fertility, pregnancy planning behaviour and patterns of access and use of reproductive health care. Linking these potential inequalities to the RSI policy permits us to evaluate its impact on health and health care access and use. Conclusions Gradients of poverty and associated patterns of behaviour and use of reproductive health care were identified. Linking these inequalities to the RSI policy also creates the opportunity of improving particular aspects of the policy itself, in order to ameliorate eventual maldistribution of health care and promote health equity. The present work is essential to subsequently model links among health determinants, health benefits/costs, key policies and economic benefits/costs.