An explosion of different Rules of Origin (ROO) has accompanied the spread of Preferential Trade Agreements (PTAs) around the world. Often tailor-made, these ROO are there to prevent trade deflection and transshipment. For developing countries, protection of regional producers of intermediate products in supply chains can be an important — if not the main — objective of PTAs, most often reciprocal Free Trade Areas (FTAs). Complying with ROO requirements entails costs. Observers, firms, customs officials and policy-makers have advocated harmonization and simplification of these ROO. The paper presents the Rule of Origin Facilitator (ROF) database developed at the International Trade Centre (ITC). Currently, the ROF contains all the texts describing ROO for 370 Preferential Trade Agreements (PTAs). These PTAs have 30 distinct Regime-wide (RW) rules and over 50,000 distinct Product Specific Rules (PSRs). Classifying PSRs by aggregating them into a small number of groups is a necessary first step prior to analysing the associated compliance costs. This paper introduces observation-based metrics to examine the extent of differences across products within a PTA and between PTAs, a first step prior to harmonization. Two indices are used to measure similarity: (i) wording using text analysis; (ii) overlap using regulatory distance. These metrics are applied to RW and PSRs for six PTAs: ASEAN, SAFTA, MERCOSUR, SADC, EUROMED and NAFTA. This choice ensures geographic coverage and is representative of the diversity of ROO around the world. The same menu of RW rules prevails across PTAs for origin provisions but greater diversity is found across the selection of certification provisions. Comparing origin and certification provisions for RW rules across the PTAs suggests that the Asian PTAs have more ‘user-friendly’ criteria. The great diversity in PSRs across products within PTAs and across PTAs presents a challenge for summarizing similarity, starting with aggregating the distinct PSRs into a small number (20) of categories of PSRs. Except for SAFTA, there is great disparity across sectors and countries over the three dimensions used to classify PSRs (frequency, coverage and distribution of preferential margins).