In 2023, the European Union set more ambitious targets for reducing greenhouse gas emissions from passenger cars: the new fleet-wide average targets became 93.6 g/km for 2025, 49.5 g/km in 2030, going to 0 in 2035. One year away from the 2025 target, this study evaluates what contribution to CO2 reduction was achieved from new conventional vehicles and how to interpret forecasts for future efficiency gains. The European Commission’s vehicle efficiency cost-curves suggest that optimal technology adoption can guarantee up to 50% CO2 reduction by 2025 for conventional vehicles. Official registration data between 2013 and 2022, however, reveal only an average 14% increase in fuel efficiency in standard combustion vehicles, although reaching almost 23% for standard hybrids. The smallest gap between certified emissions and best-case scenarios is of 14 g/km, suggesting that some manufacturers’ declared values are approaching the optimum. Yet, the majority of vehicles do not appear to fully exploit the potential of the technological boundary. In 2022, gasoline vehicles’ mass, engine size and power alone explained 67% of CO2 variation, an increase of almost 20% from 2014. For diesels, wheelbase – a proxy for vehicle size – increased in explanatory power from 5% to 18%, to the detriment of engine size, which lost 6% variance points. Vehicle mass, power, capacity and size explain well the gap between current CO2 emissions and optimal targets and may add or subtract efficiency from other energy-saving technologies. These patterns should be read in combination with the evolution of the different vehicle segments’ market shares, which saw a 40% increase in Sport Utility Vehicles (SUVs), and a sharp decrease in diesel registrations. Finally, this paper offers a statistical analysis first attempt at disentangling over time changes in vehicle characteristics from actual improvements in vehicle efficiency.