In 2011, the European Commission introduced a limit for nonvolatile particle number (PN) emissions >23 nm from light-duty (LD) vehicles and the stated intent is to implement similar legislation for on-road heavy-duty (HD) engines at the next legislative stage. This paper reviews the recent literature regarding the operation-dependent emission of PN from LD vehicles and HD engines, and the measurement procedure used for regulatory purposes. The repeatability of the PN method is of the order of 5% and higher scatter of the results can easily be explained by the effect of the vehicles or the aftertreatment devices on the PN emissions (e.g., the fill state of the diesel particulate filters). Reproducibility remains an issue since it may exceed 30%. These high-variability levels are mainly associated with calibration uncertainties of the PN instruments. Correlation measurements between the full-flow dilution tunnels (constant-volume samplers, CVS) and the proportional partial-flow dilution systems (PFDS) showed agreement within 15% for the PN method down to 1 × 10¹¹ p/kWh. At lower concentrations, the PN background of the CVS and/or the PFDS can result in larger inconsistencies. The filter-based particulate matter (PM) mass and the PN emissions correlate well down to 1–2 mg/km for LD vehicles and to 2–3 mg/kWh for HD applications. The correlation improves when only elemental carbon mass is considered: it is relatively good down to 0.1–0.3 mg/km or mg/kWh. Copyright 2012 American Association for Aerosol Research