A case of laboratory-acquired infection with Escherichia coli 0157:H7
is presented. Evidence of the identity of the infecting strain was pro
vided by toxin type and plasmid profiles. Because no obvious technical
errors in laboratory practices could be demonstrated we conclude that
the infecting dose for E. coli 0157: H7 may be small. The clinical co
urse was uncomplicated; during reconvalescence, the patient's serum re
cognized a uniqe 87 kDa band on immunoblots of the infecting strain.