The UK funding councils had originally proposed that the threshold for submitting fewer than the standard four outputs should be 14 months of absence during the assessment period, which runs from the start of 2008 to the end of 2013.
But an alternative proposal to permit women to submit one fewer output for each period of pregnancy was supported by an "overwhelming majority" of respondents to the recent consultation on the draft panel criteria for the REF.
The funding councils were not due to give their final decisions on the criteria until January, but they have announced an early decision on maternity owing to the strength of feeling. The remaining decisions will be announced in January as planned.
Opponents of the 14-month proposal were united in praise for the decision. Trevor McMillan, pro vice-chancellor for research at Lancaster University and chair of the 1994 Group's research and enterprise policy group, said the adopted proposal was "vastly fairer", taking "appropriate consideration of maternity leave without adding complexity".
Cary Cooper, distinguished professor of organisational psychology and health at Lancaster University and chair of the Academy of Social Sciences, said it was the "wholly liberal and right thing to do", bringing the academy into line with the trend in other employment sectors for improved maternity provision.
He said women would still feel pressure to submit the full four outputs, but added that "for someone who has health problems as a result of pregnancy or has a child that is ill, it allows them the space [to submit fewer] in case they need it".
Ann Blair, chair of the University and College Union's women members' standing committee, said the final guidance should address such issues, including post-natal depression, explicitly. "Clearly, these are not areas where one size will ever fit all and assessment leaders across the country must operate to the same standards," she said.