This report is the third in a series on phonological impairment in children who stutter, comparing its extent in those whose stuttering will be persistent with those in whom that disorder will disappear spontaneously. The first (E. P. Paden & E. Yairi, 1996) compared small groups of these children with normally fluent children of the same ages and sex. The second (E. P. Paden, E. Yairi, & N. G. Ambrose, 1999) compared the phonological abilities evidenced soon after onset of stuttering for 84 young children. In that study, the mean level of phonological skills of the 22 participants whose stuttering eventually persisted for at least 4 years was found to be significantly poorer than that of 62 others whose stuttering would disappear without fluency intervention before that time. In the present study, recorded performances of the same 84 children, made 1 and 2 years later, were similarly evaluated to determine how their phonological development progressed after the initial visit. Results of assessment at the 1-year visit showed that the mean difference between the two groups of children was no longer significant. The children whose stuttering would persist had improved more phonologically than had those who would recover from stuttering. At the 2-year visit, the mean percentage of phonological error for the two groups was identical. Furthermore, at this assessment, only 3 of the children in the Persistent group and 11 of those in the Recovered group had not essentially mastered all of the 10 basic patterns of phonology that were the focus of our evaluation. The findings concerning the longitudinal covariance of stuttering and phonological skills provide information that should be considered in any attempt to explain the relation between the two.