Sir.—I read with interest the article "A Half Century of Neonatal Sepsis at Yale: 1928 to 1978" by Dr Freedman and his colleagues in the Journal (1981;135:140-144). I was impressed by the relatively large number of Haemophilus influenzae isolates in the group that accounted for 2.3% of cases in the latter study group. Interestingly, six of the 11 strains were nontypable. This is in agreement with a finding by Dr McIntosh and me that was previously published.1 In fact, to my knowledge, we were the first to point out the emergence of H influenzae as an important pathogen in neonatal sepsis. Furthermore, we pointed out, at the time, the relationship of the maternal genital carriage of H influenzae to the occurrence of disease in the newborn. We screened 121 women for H influenzae in the genital tract and were able to find nontypable H influenzae in less than