The global impact of shipping on atmospheric chemistry and radiative forcing, as well as the associated uncertainties, have been quantified using an ensemble of ten state-of-the-art atmospheric chemistry models and a pre-defined set of emission data. The analysis is performed for present-day conditions (year 2000) and for two future ship emission scenarios. In one scenario emissions stabilize at 2000 levels; in the other emissions increase with a constant annual growth rate of 2.2% up to 2030 (termed the "Constant Growth Scenario''). The first key question addressed by this study is how NO_x and SO₂ emissions from international shipping might influence atmospheric chemistry in the next three decades if these emissions increase unabated. The models show future increases in NO_x and ozone burden which scale almost linearly with increases in NO_x emission totals. For the same ship emission totals but higher emissions from other sources a slightly smaller response is found. The most pronounced changes in annual mean tropospheric NO₂ and sulphate columns are simulated over the Baltic and North Seas; other significant changes occur over the North Atlantic, the Gulf of Mexico and along the main shipping lane from Europe to Asia, across the Red and Arabian Seas. Maximum contributions from shipping to annual mean near-surface ozone are found over the Atlantic (5–6 ppbv in 2000 reaching up to 8ppbv in the 2030 Constant Growth Scenario). Large increases in tropospheric ozone column are found over the Atlantic and even stronger over the Indian Ocean (1 DU in 2000 and up to 1.8 DU in 2030). Tropospheric ozone forcings due to shipping are 9.8±2.0 mW/m² in 2000 and 13.6±2.3 mW/m² in 2030. Whilst increasing ozone, ship NO_x simultaneously enhances OH, reducing the CH₄ lifetime by 0.13 yr in 2000, and by up to 0.17 yr in 2030, introducing a negative radiative forcing. Over Europe, the increase in ship emissions under the "Constant Growth Scenario'' will enhance the positive trend in NO₂ over land up to 2030. In addition, efforts to lower European sulphate levels through reductions in SO₂ emissions from anthropogenic sources on land will be partly counteracted by the rise in ship emissions. Globally, shipping contributes with 3% to increases in ozone burden until 2030 and with 4.5% to increases in sulphate. The results discussed above are calculated under the assumption that all other emissions follow the IPCC SRES A2 scenario. However, if future ground based emissions follow a more stringent scenario, the relative importance of ship emissions becomes larger. The second key issue of this work is to examine the range of results given by the individual models compared to the ensemble mean. Uncertainties in the different model approaches in the simulated ozone contributions from ships are found to be significantly smaller than estimated uncertainties stemming from the ship emission inventory, mainly the ship emission totals, the neglect of ship plume dispersion, and the distribution of the emissions over the globe.