SUMMARY Spiral filaments seen in the blood cultures of two patients with fever and jaundice were initially thought to be leptospires; these were later proved to be artefacts. An investigation was carried out to exclude the possibility of laboratory contamination of the culture media and to find out how these objects were produced. The significance of the findings is discussed in relation to the possibility of a mistaken diagnosis in routine laboratories which have a limited experience of leptospires.Two patients with fever, headache, and jaundice were admitted to two hospitals of this district within a period of six months. The first patient was a male sewer worker and the second was a female laboratory assistant who was responsible for feeding rats. Blood culture and other tests, including those for Australia antigen, urea level, microscopy and cultures of urine and faeces, and blood counts, failed to establish a diagnosis in either patient. However, fluid from blood culture bottles of both patients, when examined under dark-ground microscopy, showed spiral objects having sluggish and sinuous movement. These were thought to be leptospires as their appearances were similar to those given in a standard textbook (Cruickshank et al., 1973), although none had hooked ends and none showed active motility. The Leptospirosis Reference Laboratory in London did not confirm our findings, and the serological tests for leptospirosis were negative. Both patients recovered within a few days without any specific treatment. Investigations were made to discover the significance and source of the spiral objects seen in the blood culture bottles in our laboratory and to exclude the possibility of contamination of the media.Material and methods (a) Ten blood culture bottles prepared in this laboratory (KM) were incubated at 370C without any addition of blood. Drops of the broth were removed andexamined under the dark-ground microscope on the 1st, 3rd, 7th, and 14th days of incubation for the presence of the spiral objects.(b) Ten blood culture bottles (KM) were inoculated with blood from hospital patients with known Received for publication 24 April 1979 conditions other than leptospirosis (for example, postoperative pyrexia, bacterial endocarditis, etc). These were then incubated and examined as in (a).(c) Two blood culture bottles (KM) were inoculated with blood from two normal healthy volunteers. These were incubated and examined as in (a).(d) Five blood culture bottles were obtained from a neighbouring hospital laboratory (NG). These were inoculated with patients' blood, as in (b), and then examined as in (a).(e) Five specimens of blood containing an anticoagulant (EDTA) were obtained from the haematology laboratory of this hospital. These were incubated at 370C without the addition of any culture medium. They were then examined in the same way as above.The KM blood culture bottles contained brain heart infusion, liquoid 0-05 %, and glucose 0-5 %, and the NG bottles contained tryptone soya broth and liquoid 0 05 %.The dark-gr...