Version IV of the ENDF/B data file was released in February 1975. The description of the formats and procedures are documented in BNL-NCS-50496
A general change in this version was in the energy range for general purpose materials, where the range covered energies from 10-5 eV up to 20 MeV.
The File 1 changes included a change in the formats for specifying radioactive decay. Section MT = 453 was changed to include only production of radioactive nuclides and section MT = 457 was added to include radioactive decay data. Section MT = 456 was added to supply data for the number of prompt neutrons per fission (νp).
The File 3 changes included the energy mesh for the total cross section must include the energy meshes for partial cross sections. Time sequential (n,2n) reactions are described by using sections MT = 6-9 and MT = 46-49. An LR flag was added to designate x in the (n,n') reactions when x is not a photon. In this case, the temperature field S (formerly T) is used to designate the Q-value or energy difference of the combined reactions. Sections MT = 718, 738, 758, 778, 798 and MT = 719, 739, 759 779 and 799 are redefined to describe continuum levels for (n, x') reactions. MT = 718 describes the (n,p'c) continuum cross sections as part of the (n,p) cross section and should be included in the total cross section. MT = 719 is used to describe a continuum cross section for exit protons, whose cross section is already represented in the total cross section by other types.
Until ENDF/B-IV, the only method available to evaluators to communicate uncertainty information was through the documentation. Francis Perey designed the first data covariance format, which was approved by CSEWG in May 1973 and revised in December 1973.Three general-purpose evaluations for C, for 14N and for 16O, were released with covariance files.
-- Taken from the ENDF/B-VII format manual, Appendix I "Historic Perspective", By Norman Holden.