The calcium isotope ratios (δ44/42Ca) in milk-consuming infants and juveniles are compared to adults at two archaeological sites to test whether milk consumption is detectable in the skeleton using calcium isotopes. At Aşıklı Höyük, δ44/42Ca is correlated with δ15N and differs significantly between juveniles and adults, suggesting that δ44/42Ca is responding to dietary milk input. However, at Christ Church, Spitalfields, no correlation between δ44/42Ca and δ15N or age is noted. At this site, δ44/42Ca of adult females is not related to the minimum number of births, and we conclude that the maternal skeleton is not sufficiently isotopically perturbed for this signal to remain throughout her lifetime. Growth may result in an increase in very young infants' δ44/42Ca, in contrast to an expected decrease due to milk consumption, complicating the interpretation of the data. We conclude that milk consumption does not always result in a change in skeletal δ44/42Ca, or is masked by other effects.