The high incidence of parenteral infection in infancy, however its r6le in the production of morbidity and mortality rates is interpreted, cannot be doubted. Many statistics are available in the literature. One of the latest surveys, for example, is that by Alexander and Eiser (1944) who found evidence of parenteral infection in 88-6 per cent. of their cases, the bulk of this infection taking the form of upper respiratory infection and otitis media. The common organisms were haemolytic streptococci, pneumococci and staphylococcus aureus. Parenteral infection is even more deadly in the premature infant whose powers of resistance are feeble. Inl Walton Hospital, Liverpool, evidence of parenteral infection has been found in one-third of premature infants coming to post-mortem examination. When penicillin became avilable for civilian use it was thought that parenteral infection in infancy might be a good field for its use and immediate steps were taken to test this possibility. A gift of a small quantity of penicillin was obtained and seven cases were treated. These are included in the series about to be described (group, A, no.