Entertainment media content is often mentioned as one of the roots of children's unhealthy food consumption. This might be due to the high quantity of unhealthy foods presented in children's media environments. However, less is known about the role of the centrality of food placement, that is, whether foods are interacted with, consumed, verbally mentioned, or appear unobtrusively. We also lack longitudinal research measuring both children's unhealthy and healthy food consumption behaviors as outcomes. The aim is to connect content analytical data based on children's actual media diet with panel data in order to explain children's food preferences. Moreover, this study not only focuses on the amount of healthy and unhealthy foods children are exposed to, but also on how these foods are presented (ie, centrally or not). Furthermore, we looked at the question of how parental coviewing can diminish (or enhance) the effects of unhealthy (or healthy) food depictions, and we measured healthy and unhealthy consumption as dependent variables. We conducted a 2-wave panel study with children and one of their parents (of 2250 parents contacted, 829 responded, for a response rate of 36.84%; 648 valid cases, ie, parent-child pairs, were used for analysis), with 6 months between the 2 panel waves. We linked the 2-wave panel data for the children and their parents to content analytical data for movies (n=113) and TV series (n=134; 3 randomly chosen episodes per TV series were used) that children were exposed to over the course of 6 months. There was no significant relationship between exposure to unhealthy food presentation and unhealthy (b=0.008; P=.07) or healthy (b=-0.003; P=.57) food consumption over time. Also, healthy food presentation was unrelated to unhealthy (b=0.009; P=.18) or healthy (b=0.000; P=.99) food consumption over time. However, there was a significant, positive interaction between unhealthy food presentation and presentation centrality on unhealthy food consumption (b=0.000; P=.03), suggesting that the effects of unhealthy food presentation rise with increasing levels of centrality. There was no interaction between unhealthy food presentation and presentation centrality on the consumption of healthy foods (b=0.000; P=.10). Also, exposure to healthy food presentation interacted with centrality (b=-0.001; P=.003). That is, when a healthy product was presented at maximum centrality, it led to less unhealthy food consumption in children. Coviewing did not interact with exposure to unhealthy foods when explaining unhealthy (b=0.003; P=.08) or healthy (b=-0.001; P=.70) food consumption. We conclude that simply presenting more healthy foods is not sufficient to combat children's unhealthy food preferences. Further regulations may be necessary with respect to representations of unhealthy foods in children's media.