SummaryEarlier investigations have shown that the passive haemagglutination reaction is a specific and reliable tool in demonstrating humoral antibodies to different bacterial species.In the first part of this survey some methodologic investigations are presented. It was found that the antibodies in the sera were partly or completely inactivated during storage at +4°C for about a month, while the activity was unchanged after at least six months at ‐20°C. The limiting concentration, necessary for sensitization of the erythrocytes used in the reaction, was found to be around 60 million bacteria for different 0‐antigens.Four different sources of antigen were used during this investigation. Originally the infecting urinary bacteria from the patients were employed. Further studies compared this source with the faecal coli flora, with Kunin's common enterobacterial antigen and with a polyvalent antigen, consisting of a mixture of the eight E. coli 0‐types most frequently encountered in urinary tract infections. The reason for these investigations were partly that some source other than urinary bacteria had to be used in the controls and in suspected cases of pyelonephritis without urinary excretion of bacteria, and partly that the preparation of individual antigens for each patient makes the method too complicated and time‐consuming for use as a clinical routine procedure.These studies revealed, that the faecal bacterial antigens gave results similar to the urinary ones in about half of the cases. The common enterobacterial antigen could not be used as a substitute for the homologous homologous antigens since practically no antibodies were observed to this antigen in the sera. The polyvalent antigen was found to be a valuable substitute for the infecting strains, as equal titres could be demonstrated in those cases, where the homologous urinary strains belonged to one or another of the types, included in the antigen pool. By means of this antigen it was also found that some patients, lacking antibody response to the bacterial strains, found in their urine, did have antibodies in elevated titres against the polyvalent antigen. This might be due to a continuous antigen stimulation from earlier infections. If this is proved, the finding indicates that determination of the coli antibody titre is a valuable tool in the detection of occult chronic pyelonephritis. However, further studies are necessary to permit definite conclusions.A material of controls was studied and it was found, that almost all children above the age of about two months have agglutinating antibodies to E. coli in their serum.A series of patients with uncomplicated first‐time urinary tract infections with and without renal involvement were followed longitudinally. The findings showed that almost all of the patients above the age of three months with pyelonephritis had a sharp increase in the antibody titre, while those without functional impairment of the kidneys had titres within the limit of the control cases. It is thus possible to differentiate between cases of pyelonephritis and cases of lower urinary tract infections by means of this method. Elevated coli antibody titres were not found in pyelonephritic patients below the age of two months. The reason for this is discussed.A comparison between the antibody titre courses in patients with and without recurrences after their first episode revealed that the initial antibody response in the former group was not less than in the latter group.The rapid decrease (within 90 days) of the antibody titres found in most of the patients after treatment of the infections indicates that the homologous antigens were effectively eliminated by the treatment.Determinations of the coli antibody titre were performed in some patients, whose clinical symptoms were suggestive of pyelonephritis, but in whom the usual urinary findings were absent or inconclusive. In eight out of thirteen of these a course of the antibody titre, similar to that observed in patients with overt pyelonephritis, was demonstrated and it is therefore regarded as most likely, that these patients were in fact suffering from pyelonephritis.Further, the clinical use of antibody titrations in different types of urinary tract infections or their sequelae are illustrated by patients with recurrent urinary tract infection, renal hypertension and asymptomatic bacteriuria. The findings in some of these patients may indicate, that determination of antibodies to E. coli is a useful instrument in the detection of some of those patients who are at risk for contracting chronic pyelonephritis.