"Our hope is that these are honest mistakes and not willful attempts to mislead," said Dr. Michael Frumovitz, a professor at the University of Texas MD Anderson Cancer Center in Houston, and lead author of one of the studies.
In a field where precision is important, "even if it's an honest mistake it's very troubling," he said.
Earlier studies have found that other specialties within medicine suffer from the same problem.
Anywhere from one to 30 percent of applications to training programs in radiology, emergency medicine, orthopedics and others include references to published research that can't be located by reviewers.
Frumovitz and his colleagues and a separate group at the University of Washington, who published their reports in Obstetrics & Gynecology, wanted to find out if the same was true in their fields.
The team led by Frumovitz collected all 258 applications to a fellowship program at MD Anderson in gynecologic oncology from 2004 to 2008.
The applicants were doctors who had completed their medical school and residency training.
Of the 148 doctors who listed that they had published research findings, 44 included a reference to a publication that Frumovitz's group could not track down.
"We all believe that applicants have the best intention and are representing themselves truthfully. But (this result) falls right in line with others who have done similar work," said Frumovitz.
The University of Washington team, led by Dr. Anne-Marie Amies Oelschlager, looked through two years' worth of applications to a in obstetrics and gynecology.
Residents are doctors who have finished medical school, and are continuing their required clinical training in a particular field.
Among the 937 applicants, 357 put down that they had at least one research study that was published or about to be published in a peer-reviewed outlet.
When Amies Oelschlager's group went to find those publications, 156 of the 1,000 publications listed turned up missing.