In the years 1987 and 1988 TNO and Akzo Nobel investigated the thermal safety of non-temperature-controlled sea transport of organic peroxides, by modeling and experimenting on several scales. A full-scale test was organised in the summer of 1989, in co-operation with the former Nedlloyd company, in which a sea container with palletised packages, filled with soybean oil as inert liquid and fully equipped with temperature measuring equipment, was transported from Rotterdam, The Netherlands, to Beira, Mozambique, and back. The trip took 2 months. To answer questions stated by the workgroup International Group of Experts on the Explosion Risk of Unstable Substances-subgroup on Energetic and Oxidising Substance (IGUS-EOS) during its meeting in Washington, DC, USA, April 2006, the original data of the sea-container trip were reviewed and applied in a new mathematical model of the sea-container filled with packages, and traveling through tropical sea areas. Analysis of the data from the full-scale test led to defining a worst combination of adverse temperature conditions: constant dead sailing of the wind (wind direction and speed equal to that of the ship), and sunny weather 12 h/day and a mean ambient temperature of 30 degrees C during 28 days. The effect of sunny weather was put into the model by assigning 24 h averaged excess temperatures to the sun-exposed surfaces. New simulations showed no runaway for the sea container, if filled with an organic peroxide having an self-accelerating decomposition temperature (SADT) of exactly 55 degrees C. The exaggerated worst case with an improbable mean ambient temperature of 33 degrees C also does not lead to a runaway. Therefore, the UN criteria for temperature control for organic peroxides are appropriate. (c) 2008 Published by Elsevier Ltd.