The US National Institutes of Health (NIH) is addressing the growing concern over the lack of reproducibility in biomedical research. Scientists and the public are worried that the system ensuring reproducibility is failing and needs restructuring. The NIH, led by Francis S. Collins and Lawrence A. Tabak, is exploring interventions to restore the self-correcting nature of preclinical research. Science is traditionally self-correcting through replication, but recent issues have compromised this. Factors include poor experimental design training, emphasis on provocative statements, and lack of reporting of basic experimental details. Other issues include the use of 'secret sauce' by some researchers, and policies that encourage publication in high-profile journals, which can lead to overvaluation of research. Preclinical research, especially using animal models, is particularly vulnerable to reproducibility issues. These can be due to differences in animal strains, lab environments, or subtle protocol changes. Overinterpretation of hypothesis-generating experiments and publication bias also contribute. The NIH is developing a training module on reproducibility and transparency, to be included in mandatory training for postdoctoral fellows. It is also testing a checklist for grant applications and evaluating the scientific premise of applications. The NIH is exploring ways to increase data transparency, such as a Data Discovery Index, and launching an online forum for open discourse. The NIH is also working with the research community, publishers, and other stakeholders to improve reproducibility. Journals are being encouraged to publish negative findings and correct earlier work. The NIH is considering changes to the biographical sketch form to emphasize the significance of research contributions. It is also examining ways to anonymize the peer-review process to reduce unconscious bias. The NIH acknowledges that efforts alone are not enough and calls for the full engagement of the biomedical research community to address the issue. The goal is to reduce the frequency and severity of reproducibility problems through systematic changes.The US National Institutes of Health (NIH) is addressing the growing concern over the lack of reproducibility in biomedical research. Scientists and the public are worried that the system ensuring reproducibility is failing and needs restructuring. The NIH, led by Francis S. Collins and Lawrence A. Tabak, is exploring interventions to restore the self-correcting nature of preclinical research. Science is traditionally self-correcting through replication, but recent issues have compromised this. Factors include poor experimental design training, emphasis on provocative statements, and lack of reporting of basic experimental details. Other issues include the use of 'secret sauce' by some researchers, and policies that encourage publication in high-profile journals, which can lead to overvaluation of research. Preclinical research, especially using animal models, is particularly vulnerable to reproducibility issues. These can be due to differences in animal strains, lab environments, or subtle protocol changes. Overinterpretation of hypothesis-generating experiments and publication bias also contribute. The NIH is developing a training module on reproducibility and transparency, to be included in mandatory training for postdoctoral fellows. It is also testing a checklist for grant applications and evaluating the scientific premise of applications. The NIH is exploring ways to increase data transparency, such as a Data Discovery Index, and launching an online forum for open discourse. The NIH is also working with the research community, publishers, and other stakeholders to improve reproducibility. Journals are being encouraged to publish negative findings and correct earlier work. The NIH is considering changes to the biographical sketch form to emphasize the significance of research contributions. It is also examining ways to anonymize the peer-review process to reduce unconscious bias. The NIH acknowledges that efforts alone are not enough and calls for the full engagement of the biomedical research community to address the issue. The goal is to reduce the frequency and severity of reproducibility problems through systematic changes.