This paper analyses the effects of access to Rural Public Works (RPW) and Public Distribution System (PDS), a public food subsidy programme, on consumption poverty, vulnerability and undernutrition in India drawing upon the large household data sets constructed by National Sample Survey (NSS) data, 50th round in 1993-1994 and 61st round in 2004-2005. Treatment-effects model and Propensity Score Matching (PSM) model are used to take account of the sample selection bias in evaluating the effects of RPW or PDS on poverty. We have found significant and negative effects of the household participation in RPW and Food for Work Programmes on poverty, undernutrition (e.g. protein) and vulnerability in 1993 and 2004. On the contrary, poverty and undernutrition are significantly higher for the households with access to PDS than those without, whilst PDS has significant effects on reducing vulnerability of households in 1993 and 2004. We also applied the pseudo panel model which confirms that PDS decreased the vulnerability based on 80% of the poverty threshold. However, state-wise results of treatment effects model show considerable diversity of policy effects among different states.