For three successive growing seasons (1999–2001), a completely randomized block design experiment was established at the surrounding area of each of four sugar beet processing plants of Hellenic Sugar Industry SA, Greece (a total of 12 experiments). Nitrogen was applied at five rates (0, 60, 120, 180 and 240 kg N ha−1) and six replications per rate. Nitrogen fertilization had site‐specific effects on quantitative (fresh root and sugar yields) and qualitative (sucrose content, K, Na, α‐amino N) traits. When data were combined over years and sites, fresh root and sugar yields were maximized at high N rates (330.75 and 295 kg N ha−1 respectively), as derived from quadratic functions fitted to data. In three trials, increased N rates had negative effects on root and sugar yield. These sites were characterized by high yield in control plots, light soil texture (sand > 50 %) and low CEC values. When data were converted into relative values (the ratio of the trait values to the control mean of each experiment), root and sugar yield was found to be maximized at higher N rates (350 and 316 kg N ha−1, respectively). Sucrose content was strongly and linearly reduced by the increased N rates when data were combined but a significant reduction with increasing N rates was found in only two sites. Non‐sugar impurities (K, Na, α‐amino N) were positively related to the increased N rates when data were combined. Sodium and α‐amino N showed to be most affected by N fertilization as positive relationships were found in six and eight of 12 locations, respectively. Increased N supply resulted in higher soil NO3‐N concentrations (0–90 cm depth) at harvest which were related with amino N contents in sugar beet roots (in 1999 and 2001).