The NIH scoring system changed in 2009. The current system uses integer values, where 1 is the best score and 9 is the worst.
It is important that you read and follow the guidelines in the scoring document  and the Program Project Review Criteria described in the FOA . The numeric scores that you give should correspond well with your written critiques.
For each individual subproject:
- Before the meeting, the assigned reviewers will each provide preliminary scores for each of the five criteria (Significance, Investigator, Innovation, Approach, and Environment) and for the overall impact of the subproject. The subproject should be considered stand-alone for this scoring.
- During the review meeting, all reviewers, whether assigned or not, will provide a final score for each subproject.
- After the meeting, you may edit your criterion scores but not the final score.
For each core, you will provide a non-numeric rating of low, moderate, or high impact. The Scientific Review Officer (SRO) may ask you to use a numeric score only for the purposes of electronic scoring.
For the overall program project, you will provide a score between 1 and 9. The scores of all reviewers are averaged and multiplied by 10 to produce a final score of between 10 and 90. The overall score should take into consideration the five criteria and synergies among the components of the program project (cores and subprojects), but the final score does not have to be an average of the scores of the subprojects.