Results obtained from previous longitudinal studies of reading difficulties indicate that reading deficits are generally stable. However, little is known about the etiology of this stability. Thus, the primary objective of this first longitudinal twin study of reading difficulties is to provide an initial assessment of genetic and environmental influences on the stability of reading deficits. Data were analyzed from a sample of 56 twin pairs, 18 identical (monozygotic, MZ) and 38 fraternal (dizygotic, DZ), in which at least one member of each pair was classified as reading-disabled in the Colorado Learning Disabilities Research Center, and on whom follow-up data were available. The twins were tested at two time points (average age of 10.3 years at initial assessment and 16.1 years at follow-up). A composite measure of reading performance (PIAT Reading Recognition, Reading Comprehension and Spelling) was highly stable, with a stability correlation of .84. Data from the initial time point were first subjected to univariate DeFries-Fulker multiple regression analysis and the resulting estimate of the heritability of the group deficit (h2g) was .84 (± .26). When the initial and follow-up data were then fitted to a bivariate extension of the basic DF model, bivariate heritability was estimated at .65, indicating that common genetic influences account for approximately 75% of the stability between reading measures at the two time points.