Proteomics approaches are being increasingly applied in ecotoxicology on the premise that the identification of specific protein expression changes in response to a particular chemical would allow elucidation of the underlying molecular pathways leading to an adverse effect. This in turn is expected to promote the development of focused testing strategies for specific groups of toxicants. Although both gel-based and gel-free global characterization techniques provide limited proteome coverage, the conclusions regarding the cellular processes affected are still being drawn based on the few changes detected. To investigate how specific the detected responses are, we analyzed a set of studies that characterized proteome alterations induced by various physiological, chemical and biological stressors in zebrafish, a popular model organism. Our analysis highlights several proteins and protein groups, including heat shock and oxidative stress defense proteins, energy metabolism enzymes and cytoskeletal proteins, to be most frequently identified as responding to diverse stressors. In contrast, other potentially more specifically responding protein groups are detected much less frequently. Thus, zebrafish proteome responses to stress reported by different studies appear to depend mostly on the level of stress rather than on the specific stressor itself. This suggests that the most broadly used current proteomics technologies do not provide sufficient proteome coverage to allow in-depth investigation of specific mechanisms of toxicant action. We suggest that the results of any differential proteomics experiment performed with zebrafish should be interpreted keeping in mind the list of the most frequent responders that we have identified. Similar reservations should apply to any other species where proteome responses are analyzed by global proteomics methods. Careful consideration of the reliability and significance of observed changes is necessary in order not to over-interpret the experimental results and to prevent the proliferation of false positive linkages between the chemical and the cellular functions it perturbs. We further discuss the implications of the identified "top lists" of frequently responding proteins and protein families, and suggest further directions for proteomics research in ecotoxicology. Apart from improving the proteome coverage, further research should focus on defining the significance of the observed stress response patterns for organism phenotypes and on searching for common upstream regulators that can be targeted by specific assays.