In this study, we used the data from the study of Kids in Taiwan: National Longitudinal Study of Child Development and Care to understand parent-child shared book reading practices in Taiwan and the association between home environment factors, such as socio-economic status and childcare operation, and shared book reading practices as well as between shared book reading practices and children’s development in cognition, language, and socio-emotions when they were 36 months old. The total number of children who participated in this study was 2,164. This study found that about two thirds of the parents read books to their children at least once a week, and one-third of the parents never read to their children or read to them less than once a week. Moreover, using structural equation modeling analysis, three latent variables were constructed, namely the socio-economic status, shared book reading practices, and the development of the child. The results indicated that socio-economic status had a significant relation with shared book reading practices. The β value was 0.6, which indicated that the socio-economic status factor could explain 36% variation of shared book reading practices. Furthermore, shared book reading practices had a β value of 0.51 (i.e., explaining 28% variation) for its relation with the development of the child. Socio-economic status, however, exhibited an insignificant relation with the development of the child. In this sense, shared book reading practices mediated the relationship between the socio-economic status and the development of the child. Shared book reading practices exhibited the highest effect on the cognitive memory development of a child and had a large effect on the language expression and emotional development of the child. In this study, we also compared different subgroups to investigate the existence of moderating factors and found that the respondent types (father or mother), child’s gender, and birth order did not have any moderating effect on the causal model. Noninvariance of structural weights was observed in three subgroups of childcare operations. A stronger relation between the socio-economic status and shared book reading practices was observed for parents who did not use any childcare services (i.e., children who were cared by their own parents full time) than for parents who used childcare services less than 10 hours per day and parents who used childcare services 11-24 hours per day. The effects of shared book reading practices on the child’s development was the largest for the group using 11-24 hours of childcare services but was the least for the group that did not use childcare services. This result may imply that shared book reading practices can improve the child’s development most effectively though parents who can only spend limited time with their children.