From 1999 to 2006, 36 field experiments were conducted in five sugar beet growing areas in Greece (Larissa, Plati, Serres, Xanthi and Orestiada) to monitor yield. Locations differed significantly regarding thermal variables during the growing season with Xanthi having the most favourable thermal conditions (Tmax, average daily maximum temperature; Tmean, average daily mean temperature; GDD, growing degree days) for sugar beet growth. From early June to the end of the harvesting campaign, successive harvests were conducted. Over the years, fresh root weight and sugar yield at the last harvest of the season (FRWLH, SYLH) did not differ significantly among locations. Also, there were no significant differences among locations regarding GDD for maximum FRW and SY (GDDMFRW, GDDMSY), with the means over location estimated at 2639·9 and 2792·5 °C, respectively. Days after seeding (DAS) necessary for maximum yield (DASMFRW, DASMSY, respectively) differed among locations, with the longest period (DASMFRW 206·4 days, DASMSY: 204·5 days) occurring in the northernmost location (Orestiada). Means for DASMFRW and DASMSY at the five locations were estimated at 190·4 and 188·9·days, respectively. Excluding Xanthi and combining the remaining locations, FRWLH and SYLH were negatively correlated with the average temperatures (Tmean, Tmax and Tmin, daily minimum temperature) over the growing season. The opposite was evident for Xanthi where sugar beet was grown under sub-optimal temperatures. The optimum mean Tmax of the five locations was estimated at 25·5 and 25·1 °C for FRWLH and SYLH, respectively. Elongation of the growing season, by means of early sowing, would increase yield by decreasing average temperatures (Tmean, Tmax) over the growing season in locations with the highest recorded temperatures (Larissa, Plati, Serres and Orestiada). In Xanthi, the projected temperature increase, as a result of climate change, is expected to have a positive effect on yields.