Recent studies on dendrochronologically-dated wood from the Southern Hemisphere show differences in radiocarbon calibration curves between the hemispheres. Therefore, it is recommended to use calibration data from securely dated wood in the corresponding hemisphere. This paper outlines recent work and recommends the use of an internationally agreed data set for future Southern Hemisphere radiocarbon dating. Hogg et al. (2002) and McCormac et al. (1998) measured radiocarbon activity in wood samples from the Northern and Southern Hemispheres, showing that a fixed offset between hemispheres is not reliable. Instead, the Southern Hemisphere calibration should use securely dated wood with replicate radiocarbon measurements. The Southern Hemisphere is defined as south of the thermal equator or Intertropical Convergence Zone (ITCZ). Two workshops in 2002 and 2003 reviewed recent developments in radiocarbon calibration. Data from the Southern Hemisphere, including measurements from New Zealand, Chile, and South Africa, were considered. It was agreed that a combined set of Southern Hemisphere measurements should form a single calibration, provided they showed good consistency. This was ratified at the International Radiocarbon Conference in Wellington, New Zealand (2003). The Southern Hemisphere calibration curve, SHCal04, was constructed using data from McCormac et al. (2002) and additional measurements from the Pretoria data set. Laboratory error multipliers were used to account for uncertainties. The calibration curve beyond 1000 cal BP was based on a random effects model, considering the varying offset between hemispheres. The offset between hemispheres is not constant but varies from 1 to 10‰. A random effects model was used to account for this variation. The offset is larger than previously published averages but reflects the trend from 50–990 cal BP. The Southern Hemisphere calibration curve extends to 1000 cal BP, with data available online. The distinction between Southern and Northern Hemisphere atmospheres should be based on the thermal equator or ITCZ, not the geographic equator. This introduces uncertainty for tropical and neo-tropical sites. The ITCZ's seasonal shift may affect radiocarbon measurements. The SHCal04 data set is recommended for future Southern Hemisphere calibration back to 11.0 cal kyr BP. Before the Holocene, larger-scale carbon reservoir changes may have altered the interhemispheric offset. The international community is encouraged to support further dendrochronological records in the Southern Hemisphere.Recent studies on dendrochronologically-dated wood from the Southern Hemisphere show differences in radiocarbon calibration curves between the hemispheres. Therefore, it is recommended to use calibration data from securely dated wood in the corresponding hemisphere. This paper outlines recent work and recommends the use of an internationally agreed data set for future Southern Hemisphere radiocarbon dating. Hogg et al. (2002) and McCormac et al. (1998) measured radiocarbon activity in wood samples from the Northern and Southern Hemispheres, showing that a fixed offset between hemispheres is not reliable. Instead, the Southern Hemisphere calibration should use securely dated wood with replicate radiocarbon measurements. The Southern Hemisphere is defined as south of the thermal equator or Intertropical Convergence Zone (ITCZ). Two workshops in 2002 and 2003 reviewed recent developments in radiocarbon calibration. Data from the Southern Hemisphere, including measurements from New Zealand, Chile, and South Africa, were considered. It was agreed that a combined set of Southern Hemisphere measurements should form a single calibration, provided they showed good consistency. This was ratified at the International Radiocarbon Conference in Wellington, New Zealand (2003). The Southern Hemisphere calibration curve, SHCal04, was constructed using data from McCormac et al. (2002) and additional measurements from the Pretoria data set. Laboratory error multipliers were used to account for uncertainties. The calibration curve beyond 1000 cal BP was based on a random effects model, considering the varying offset between hemispheres. The offset between hemispheres is not constant but varies from 1 to 10‰. A random effects model was used to account for this variation. The offset is larger than previously published averages but reflects the trend from 50–990 cal BP. The Southern Hemisphere calibration curve extends to 1000 cal BP, with data available online. The distinction between Southern and Northern Hemisphere atmospheres should be based on the thermal equator or ITCZ, not the geographic equator. This introduces uncertainty for tropical and neo-tropical sites. The ITCZ's seasonal shift may affect radiocarbon measurements. The SHCal04 data set is recommended for future Southern Hemisphere calibration back to 11.0 cal kyr BP. Before the Holocene, larger-scale carbon reservoir changes may have altered the interhemispheric offset. The international community is encouraged to support further dendrochronological records in the Southern Hemisphere.