C. Sinha and B. Carabine's (1981, Journal of Child Language, 8, 109–129) claim of a U-shaped course to the development of quantitative knowledge with an early stage of conservation based on functional significance was tested in two experiments. Children at three age groups from 3 to 6 years were presented with the original procedure or control versions in which the display was slightly altered. Results showed that although Sinha and Carabine's result was replicable, their interpretation is unwarranted. Performance in the younger groups was best interpreted as the use of nonquantitative possession or position strategies.