It is generally assumed that one of the reasons why diabetics are more susceptible to urinary tract infections than non-diabetics is their 'sweet urine'. However, very little information is available on this subject. Therefore, the growth rates of different Escherichia coli strains were studied in human urine with and without added glucose and with and without a constant pH, and compared with their growth rates in Mueller-Hinton broth (MHB). Eight isolates were used (three from blood cultures from urosepsis patients, two urinary isolates, two faecal isolates and one laboratory strain K12). All isolates grew better in MHB than in urine, but with the exception of the laboratory strain, they had the same growth rate in urine. No significant difference was found between the growth rate in urine from diabetics without glucosuria and that in urine from non-diabetics. The addition of glucose (up to a concentration of 1000 mg/dl) to urine and MHB enhanced the growth rate of all isolates. However, very high concentrations of glucose (up to 10000 mg/dl) in urine and MHB caused a decrease in bacterial growth rate when the urinary pH was not kept constant. The stationary phase was reached later and the final bacterial yield was greater when the urine was made less acidic. As the uropathogenic strains did not grow better in urine than the other isolates, it may be concluded that better growth in urine is not one of the causes of the greater virulence of these strains.