An analysis of a 20-yr dataset of tropical cyclones in the southwest Pacific has been made to determine the spatial patterns of cyclone occurrence and how these depend on the Southern Oscillation (SO) and sea surface temperature. A local measure of cyclone incidence is defined as the number of cyclones that enter into a 2° lat-long square during some period (eg., within one season) and this quantity is mapped for four Southern Oscillation index (SOI) categories. The maps show that the geographical distribution of cyclone incidence shifts eastward and northward during negative SOI phases and vice versa. The mean annual incidence for the whole region is 28% higher than average for our strongly negative SO category and 16% below average for our near-zero category. To explore the roles of the SO and local SST, correlation analysis was applied to mean values of cyclone incidence for four subregions of 20° longitude width between 10° and 22°S. Markedly different SOI/SST responses were found on either s... Abstract An analysis of a 20-yr dataset of tropical cyclones in the southwest Pacific has been made to determine the spatial patterns of cyclone occurrence and how these depend on the Southern Oscillation (SO) and sea surface temperature. A local measure of cyclone incidence is defined as the number of cyclones that enter into a 2° lat-long square during some period (eg., within one season) and this quantity is mapped for four Southern Oscillation index (SOI) categories. The maps show that the geographical distribution of cyclone incidence shifts eastward and northward during negative SOI phases and vice versa. The mean annual incidence for the whole region is 28% higher than average for our strongly negative SO category and 16% below average for our near-zero category. To explore the roles of the SO and local SST, correlation analysis was applied to mean values of cyclone incidence for four subregions of 20° longitude width between 10° and 22°S. Markedly different SOI/SST responses were found on either s...