Although experts agree that Web-based health information often contains exaggeration and misrepresentation of science, it is not yet known how this information affects the readers' sentiments. This study aimed to investigate whether specific aspects of Web-based diabetes research news are associated with positive or negative sentiments in readers. A retrospective observational study of the comments on diabetes research news posted on Facebook pages was conducted as a function of the innovations' developmental phase, the intended treatment effect, and the use of strong language to intensify the news messages (superlatives). Data for the investigation were drawn from the diabetes research news posted between January 2014 and January 2018 on the two largest Dutch Facebook pages on diabetes and the corresponding reader comments. By manually coding these Facebook user comments, three binary outcome variables were created, reflecting the presence of a positive sentiment, the presence of a negative sentiment, and the presence of a statement expressing hopefulness. Facebook users made a total of 3710 comments on 173 diabetes research news posts that were eligible for further analysis. Facebook user comments on posts about diabetes prevention (odds ratio [OR] 0.55, 95% CI 0.37-0.84), improved blood glucose regulation (OR 0.68, 95% CI 0.56-0.84), and symptom relief (OR 0.31, 95% CI 0.21-0.44) were associated with less positive sentiments as compared with potential diabetes cures. Furthermore, comments on innovations supported by preclinical evidence in animals were associated with more positive sentiments (OR 1.46, 95% CI 1.07-1.99) and statements expressing hope (OR 1.47, 95% CI 1.01-2.14), when compared with innovations that have evidence from large human trials. This study found no evidence for the associations between language intensification of the news posts and the readers' sentiments. Our finding that the attitudes toward diabetes research news on Facebook are most positive when clinical efficacy is not (or not yet) proven in large patient trials suggests that news authors and editors, as well as medical professionals, must exercise caution when acting as a conduit for diabetes research news. ©Hans Vehof, Eibert Heerdink, José Sanders, Enny Das. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 13.11.2019.