Latent growth curve models (LGCM) has emerged as a flexible technique for examining longitudinal changes over time. In this study, we use data from a community-based intervention program, the Stop Now and Plan (SNAP), as an example to illustrate how LGCMs can be used in program evaluation, what one should take into consideration, and how the use of LGCM can lead to a more complete understanding of the intervention program. Univariate LGCMs were used to examine changes in children’s behaviour problems and prosocial skills and changes in parenting self-efficacy through the program. Multivariate LGCMs were used to examine associations between children’s behaviour problems and parenting self-efficacy. We will discuss statistical assumptions, missing data, strengths and limitations in use of LGCM for longitudinal data analysis and program evaluation.