Abstract-One of the hardest tasks of a certification infrastructure is to manage revocation. This process consists in collecting and making the revocation status of certificates available to users. Research on this topic has focused on the trade-offs that different revocation mechanisms offer. Much less effort has been conducted to understand and model real-world revocation processes. For this reason, in this paper, we present a novel analysis of realworld collected revocation data and we propose a revocation prediction model. The model uses an Autoregressive Integrated Moving Average Model (ARIMA). Our prediction model enables certification authorities to forecast the percentage of revoked certificates.