Resilience is, nowadays, one of the keywords in the policy debate on development. Measuring resilience and how it varies over time is dramatically important for policy makers and people living in risk-prone environments. This paper applies econometric techniques for estimating household resilience using the so-called Resilience Index Measurement and Analysis (RIMA) recently proposed by FAO (2013). It then adopts transition matrices to estimate how resilience changes over time. Finally, multinomial logit and bivariate probit models are estimated to identify the main drivers of change.