SUMMARYIn most regions, sugar beet is normally sown as a spring crop. If sown in autumn the crop remains on the field over winter and may achieve fast re-growth in spring from assimilates stored within the beet, allowing earlier leaf growth and light interception in spring. The specific leaf area (SLA) (ratio between leaf surface and leaf mass) is mainly affected by leaf area expansion and consequently affects productivity in early growth stages. The aim of the present study was (i) to examine the SLA dynamics of autumn-sown sugar beet before and after winter and (ii) to develop an empiric approach describing SLA changes during the growth period. A field trial in northern Germany with three different sowing times (mid-April, mid-June and mid-August) and varying plant densities (148 000, 246 000 and 370 000 plants/ha) was carried out in 2009/10 to 2011/12. The average SLA of the canopy was the highest (>25 m²/kg) directly after emergence, then decreased until autumn (<13 m²/kg) and increased again up to 20 m²/kg during re-growth of winter sugar beet in spring. A stepwise multiple regression analysis revealed mean photosynthetically active radiation over 10 days before measurement (PAR_mean), leaf area index (LAI), mean temperature over 10 days before measurement (T_mean) and temperature sum since sowing (T_sum) as the main influences on SLA dynamics. The strongest correlation to SLA was shown by T_mean (r = 0·69) and the weakest by T_sum (r = −0·28). A multiple linear regression model was fitted to the dataset with T_mean, PAR_mean and log (T_sum) achieving an adjusted R² of 0·64. This empirical equation is suitable for use in a crop growth model for winter sugar beet.