<jats:p>Abstract. The interaction between iron availability and the phytoplankton elemental composition was investigated during the in situ iron fertilization experiment EIFEX and in laboratory experiments with the Southern Ocean diatom species Fragilariopsis kerguelensis and Chaetoceros dichaeta. Contrary to other in situ iron fertilization experiments we observed an increase in the BSi:POC, BSi:PON, and BSi:POP ratios within the iron fertilized patch during EIFEX. This is possibly caused by a relatively stronger increase in diatom abundance compared to other phytoplankton groups and does not necessarily represent the amount of silicification of single diatom cells. In laboratory experiments with F. kerguelensis and C. dichaeta no changes in the POC:PON, PON:POP, and POC:POP ratios were found with changing iron availability in both species. BSi:POC, BSi:PON, and BSi:POP ratios were significantly lower in the high iron treatments compared to the controls. In F. kerguelensis this was caused by a decrease in cellular BSi concentrations and therefore possibly less silicification. In C. dichaeta no change in cellular BSi concentration was found. Here lower BSi:POC, BSi:PON, and BSi:POP ratios were caused by an increase in cellular C, N, and P under high iron conditions. These results indicate that iron limitation does not always increase silicification in diatoms and that changes in the BSi:POC, BSi:PON, and BSi:POP ratios under iron fertilization in the field are caused by a variety of different mechanisms. Our results therefore imply that simple cause-and-effect relationships are not always applicable for modeling of elemental ratios. </jats:p>