This paper describes a recommendation system for child care blogs based on the similarity of multidimensional sequences representing temporal child growth data. The number of users who want to read child care blogs and exchange information on child care problems is increasing. However, it is difficult for users to search child care blogs which provide suitable information for their children due to the large number of child care blogs. The use of a multidimensional sequence search method was proposed to identify sequences similar to a child's temporal growth data published in child care blogs. This allowed users to easily identify suitable child care blogs. At last, a prototype system is implemented and is evaluated.