This study aims to classify latent profiles based on parental parenting stress, identify differences in child abuse across latent profiles, and analyze the predictive influence of various factors on the classification of latent profiles. The study participants comprised 1,140 children and their parents who responded to the 14th wave of the Korean Children’s Panel Survey. The study methods included latent profile analysis(LPA), chi-square tests, and multinomial logistic regression analysis. The results of the study are as follows: First, five latent profiles of parenting stress were identified: 'Low-level Parenting Stress Profile,' 'Moderate-level Parenting Stress Profile,' 'High-level Parenting Stress Profile,' 'Father-dominant Parenting Stress Profile,' and 'Mother-dominant Parenting Stress Profile.' Second, regarding differences in child abuse across latent profiles, the 'Low-level Parenting Stress Profile' showed the lowest levels of child abuse, making it a low-risk profile for child abuse. In contrast, the 'High-level Parenting Stress Profile' and the 'Father-dominant Parenting Stress Profile' exhibited the highest levels of child abuse, categorizing them as high-risk profiles. Third, parental depression, subjective happiness, and marital satisfaction significantly predicted the classification of parenting stress latent profiles, with the influence of each factor varying by latent profile. This study classified parenting stress latent profiles and examined differences in child abuse across these profiles, confirming the relationship between parenting stress and child abuse. Furthermore, the findings highlight implications for tailored intervention strategies to prevent child abuse based on latent profile characteristics.