Children's spelling development is often described by researchers and
educators as proceeding through a series of stages. Two properties of
stages were analyzed in this study. If spelling development can be cha
racterized by stages, then it should be possible to observe qualitativ
ely different spellings at different points in development. In additio
n, spellings within a point of development must be consistent. Spellin
g samples were obtained from stories written by children in first thro
ugh sixth grade. Stage classifications of spellings for (a) silent -e
long vowel words (e.g., bake), and (b) regularly affixed past tense wo
rds phonologically represented as /t/(e.g., helped), /d/(e.g., opened)
, and /ad/(e.g., listed) were analyzed. Little evidence was found for
either predicted qualitative differences in stage classification of er
rors or in stage constancy across grades. Implications for theories of
spelling development and instructional practice are discussed.