Fish domestication progresses through five levels: from the initial acclimatization to captivity (Level 1), to the life cycle completion in captivity (Level 4), and even to the implementation of selective breeding programs (Level 5). Domestication leads to phenotypic changes over generations, sometimes from the very first generation. Behavioral traits are among the first to change. However, in fish, potential behavioral changes during early domestication have been little studied. Therefore, we studied potential behavioral changes among early and advanced levels of domestication in a model species, the zebrafish (Danio rerio), using a mirror test experiment, commonly used to assess traits involved in activity, aggressiveness, and stress in this species. We compared these traits between wild zebrafish in captivity (F0; Level 1), the first generation of their captive-born offspring (F1; Level 4), and three laboratory strains (AB, TU, and WIK; Level 5). Each fish was individually filmed and tracked using an automated procedure for 5 min. Nine behavioral traits and one activity-related trait were characterized for each individual based on the movements and positioning of the fish. We applied a principal component analysis (PCA) and tested the significance of potential differences between groups using an analysis of similarities (ANOSIM). We applied an indicator value analysis (IndVal) to determine which traits were most expressed by each group. We detected differences between groups and across domestication levels. More specifically, we highlighted differentiations between different levels of domestication (e.g. between F1, AB, TU, and WIK) as early as the beginning of the domestication process (i.e. F0 vs. F1), but also within the same level of domestication (i.e. AB vs. TU). Based on PCA and IndVal, (i) F0 and F1 tended to show stronger expression of stress-related traits than the other groups, (ii) F0 was more active than others, and (iii) TU was more aggressive than AB. Our results confirmed that domestication can change fish behavior, even in the first generation born in captivity, although these modifications remain limited. In contrast, we did not observe any general trends correlated with domestication levels, given that AB and TU diverged in their aggressiveness levels, and WIK differed only from F1. This result needs to be generalized to other species but also considered for domestication for aquaculture. If future studies confirm that the changes observed at the beginning of the domestication process remain limited and that there is no consistent evolutionary trend across generations in fish, this would highlight the crucial importance of selecting the right species from the outset of domestication. It would also emphasize the need to design selective breeding programs that shape fish stocks with the most desirable characteristics. To our knowledge, this study is one of the few to examine the behavior of wild zebrafish alongside laboratory strains, offering a unique insight into the early stages of domestication.