The persistence of academic difficulties from childhood through adulthood has led researchers to focus on the identification of the early factors influencing children's subsequent achievement in order to improve the efficient screening of children who might be at risk of school failure. The foundations of academic achievement can be accurately traced back to the preschool years prior to children's entry in formal schooling and are largely influenced by environmental determinants. Importantly, some environmental conditions act as early risk factors undermining children's later academic achievement due to the well-established relation between underachievement and exposure to moderate to high levels of environmental risk. In the present study, we aimed to investigate the longitudinal effects of environment-level factors (sociodemographic and family characteristics) and early risk exposure at kindergarten on children's subsequent academic achievement at the end of middle school (grade 9). The sample of analysis comprised 654 kindergarteners aged 5-6 years (2001-2002 school year) followed through the end of middle school when they were aged 14-15 years (2010-2011 school year). At kindergarten, assessment included questionnaire-based measures of sociodemographic and family background characteristics. These included an original set of information pertaining to family background including parental nationality, education level, history of reading difficulties, type of early childcare, family situation, family size, and language-based bedtime routines, as well as individual-level factors such as children's first language, medical history, language delay, birth weight, age of walking onset, and gestation period. At grade 9, outcome measures were composed of children's results in the national evaluations performed at the end of middle school ("Diplôme National du Brevet"), or history of repetition for a second year of the same class. The results indicated that all family background characteristics at kindergarten were related to later academic outcomes at grade 9. From the original set of family characteristics, parental educational level, family situation, language-based bedtime routines, and type of early childcare significantly predicted later academic achievement at grade 9. Moreover, a multiple risk index score aggregating these specific family characteristics, together with three individual-level factors (gender, medical history, and language delay) was robustly and positively associated with an increased likelihood of school failure at the end of middle school. Unique to our study was the finding relative to the longitudinal association over a 10-year span of language-based bedtime routines with children's academic performance at the end of middle school. These findings underline the importance of including family background information in early surveillance procedures in order to improve the efficient screening of children who might be at risk of academic underachievement. Importantly, some of these contextual factors represent environmental characteristics that can be reversed early in life through appropriate and informed support to families. Moreover, the present work has important implications regarding the early detection of children who are at familial risk of underachievement, allowing the activation and promotion of adequate intervention strategies early in children's educational trajectories.