The overall experience with latex-modified concrete (LMC) overlays constructed with Types I-II, III, and K cements and rapid set cement continues to be good. The most frequent problem with LMC since the late 1990s has been the high incidence of LMC samples failing the permeability test (AASHTO T277) when done at an early age. Approximately one-third of the field-prepared specimens fail because they have permeability values greater than 1,500 coulombs. Research was done to determine the cause and significance of the failing early-age LMC permeability test results and to recommend changes in practice. The approach included testing of specimens prepared in the laboratory, specimens prepared in the field, and cores from three bridge overlays that had failing early-age test results for field-prepared specimens. Results indicate that the permeability of LMC decreases as the concrete ages. LMC that meets the specification with the exception of early-age permeability should have a very low permeability at later age regardless of the early-age permeability. The permeability test conducted on samples of LMC at early age does not provide a good indication of the later-age permeability of LMC and should not be used to reject LMC.