From the early 19th century to the late 20th century, the mean stature of Swedish conscripts increased from 167 to 177 cm. Based on osteological data, estimated statures of medieval skeletal populations indicate that mean stature of males had decreased from an average 173 cm during the medieval period, to 165 cm at the mid 19th century. This change with respect to stature could possibly be linked to the change in diet. Based on the skeletal material unearthed from Linköping, Sweden, we explored the possible association between a changing diet and a decreasing stature. We compared the medieval sample (1100–1300 AD) to a sample dated 1780–1810 AD. A significant multiple regression analysis demonstrates that both sex and chronology explain the variation seen in femur length, but not stable isotopes of carbon and nitrogen. The Bayesian mixing model with uninformative priors suggests that pike is dominating in both the medieval and historical samples, followed by pig, cattle, cod and herring. Using an informative prior based on the consumption from the 16th century Swedish warship Lindormen did not alter the consumption pattern shown by the inhabitants of Linköping. As our results do not suggest a major shift in diet, as well as pointing out lacustrine fish as well as pig as more important to the protein in the diet than cattle, it deviates from what is inferred based on historical records. We conclude that the diet in a town cannot be generalized from the consumption of a ship and probably not from institutions either. These contexts have dominated the reconstructions of 19th century diet in Sweden based on historical records. However, there is also the possibility that the two isotopes employed are not sufficiently delicate to detect such dietary changes. Copyright © 2012 John Wiley & Sons, Ltd.