Cruise ships are making improvements in carbon efficiency.

To estimate carbon emissions from the cruise ship industry, self-reported emissions from the two main companies were used. Both Royal Caribbean Cruises Ltd and Carnival Corporation & PLC publish sustainability reports that detail annual emissions. The third largest company, Norwegian Cruise Line Holdings Ltd, does not report carbon emissions. Using the two existing sustainability reports, and in combination with the known market share reported on Cruise Market Watch, the remaining 28.8% of the global cruise activity were extrapolated.


  • Royal Caribbean Cruises Ltd and Carnival Corporation & PLC make up 71.2% of global cruise activity measured in passenger numbers.
  • The latest available data are from 2016. Carnival reported emission totals of 10,539 kilo-tonnes of carbon dioxide equivalent, a 1.6% increase compared with 2015. This corresponds to a market share in terms of passenger numbers of 48.1%.
  • Royal Caribbean Cruises Ltd emitted a total of 4,465 kilo-tonnes of carbon dioxide equivalent; a marginal increased compared with the previous year. Royal Caribbean’s market share was 23.1%.
  • The third largest company, Norwegian Cruise Line Holdings Ltd do not publicly report total carbon emissions, and all other companies have very small market shares.
  • Based on an estimated total number of about 23 million cruise ship passengers in 2015 (see de Grosbois, 2016), it can be estimated that the average cruise ship passenger emits 0.83 tonnes of carbon dioxide-equivalent for their cruise. This is equivalent to a return air trip from London to Tokyo in economy class.
  • CruiseMarket Watch reports an annual growth of 6.55%.
  • Cruise ship emissions make up 0.2% of all global carbon dioxide emissions from fossil fuel combustion and cement production. Note that tourist water transport is much larger than the specific segment of cruise ship tourism analysed here.
  • When examining the cruise ship company reports it emerges that every company uses slightly different metrics. This poses major problems in comparing and compiling data, for example some indicators use resource use per kilometer and others refer to person-day or berth.