Comparisons of the developmental pathways of the first 5 years of life for children with (N = 107) and without (N = 93) familial risk for dyslexia observed in the Jyväskylä Longitudinal study of Dyslexia are reviewed. The earliest differences between groups were found at the ages of a few days and at 6 months in brain event-related potential responses to speech sounds and in head-turn responses (at 6 months), conditioned to reflect categorical perception of speech stimuli. The development of vocalization and motor behavior, based on parental report of the time of reaching significant milestones, or the growth of vocabulary (using the MacArthur Communicative Development Inventories) failed to reveal differences before age 2. Similarly, no group differences were found in cognitive and language development assessed by the Bayley Scales of Infant Development and the Reynell Developmental Language Scales before age 2.5. The earliest language measure that showed lower scores among the at-risk group was maximum sentence length at age 2. Early gross motor development had higher correlation to later language skills among the at-risk group rather than the control children. The most consistent predictor of differential development between groups was the onset of talking. Children who were identified as late talkers at age 2 were still delayed at the age 3.5 in most features of language-related skills-but only if they belonged to the group at familial risk for dyslexia. Several phonological and naming measures known to correlate with reading from preschool age differentiated the groups consistently from age 3.5. Our findings imply that a marked proportion of children at familial risk for dyslexia follow atypical neurodevelopmental paths. The signs listed previously comprise a pool of candidates for early predictors and precursors of dyslexia, which await validation.