BackgroundIn China, the figure for left-behind children (LBC) of migrants stood at 68. 77 million in 2015. Despite being seen as a whole in the last few decades, LBC today differ broadly in parental migrating status. This study focused on LBC with both parents migrating (BLBC), LBC with only mothers migrating (MLBC), LBC with only fathers migrating (FLBC), and previous LBC with one or both parents migrating (PLBC), separately. We aimed at exploring the extent to which LBC were being affected by each migrant parent on both mental health and risk behaviors.MethodsData from 4,832 children were collected by a school-based survey in both rural and urban areas of China's Anhui province. Each participant anonymously completed a self-administered questionnaire containing the sociodemographics, the Strength and Difficulties Questionnaire (SDQ), the items from the Youth Risk Behavior Surveillance System (YRBSS), and Young's Internet Addiction Test for Chinese (YIAT-C). Data were analyzed using one-way ANOVA and the Chi-squared test. Associations were estimated by multiple linear regression and logistic regression analyses adjusted for several confounders.ResultsThe results suggested that BLBC (p < 0.001), MLBC (p < 0.05), FLBC (p < 0.01), and PLBC (p < 0.001) significantly scored higher for total difficulties along with emotional symptoms and conduct problems than never left-behind children (NLBC). Besides, BLBC, FLBC, and PLBC further reported a significantly higher rate of smoking (p < 0.001, p < 0.01, and p < 0.001, respectively) and drinking (p < 0.01, p < 0.05, and p < 0.01, respectively) than did NLBC. Also, MLBC appeared higher risks of smoking problems [OR = 2.31, 95% CI (1.45–3.69), p < 0.001] and the internet addiction [OR = 2.15, 95% CI (1.24–3.72), p < 0.01], when compared to NLBC.ConclusionsThe findings provided insight into LBC within the different contexts of parental migrations and contributed to a better understanding of their specific and potentially persistent health risks. Correspondingly, the study highlighted the implications for differentiating LBC to capture the more vulnerable group and tailored interventions to prioritize.