Antibody-coated bacteria tests, using fluorescein-conjugated rabbit anti-canine immunoglobulin (Ig)G and anti-IgA, were conducted on 77 specimens of bladder urine obtained by antepubic cystocentesis from 25 dogs with urinary tract infections. Urine specimens from 3 dogs infected with Staphylococcus aureus were tested a total of 27 times, specimens from 10 dogs infected with Escherichia coli were tested 25 times, urine from 6 dogs infected with Proteus mirabilis was sampled and tested 18 times, urine from 3 dogs with Klebsiella pneumoniae was sampled and tested 4 times, and specimens from 3 dogs infected with P rettgeri, Enterobacter cloacae, or Streptococcus faecalis were tested once each. Specimens of urine from the renal pelves of each dogs were obtained by percutaneous nephropyelocentesis at the same time as the specimens of bladder urine. The results of bacterial cultures of these specimens were compared with the results of the corresponding antibody-coated bacteria tests. False-positive results (ie, evidence of antibody coating of bladder urine bacteria but no bacterial growth from corresponding renal pelvis urine specimens) were obtained in 13 of 22 (59%) tests of dogs infected with S aureus, 12 of 13 (92%) tests in dogs infected with P mirabiis, 5 of 18(28%) tests in dogs infected with E coli, and 1 of 4 (25%) tests in dogs infected with K pneumoniae. False-negative results (ie, bacterial growth from renal pelvis urine specimens but no evidence of antibody coating of bacteria in the corresponding specimen of bladder urine) occurred in 5 of 7 (71%) specimens infected with E coli, in 4 of 5 (80%) specimens infected with P mirabilis, and in 5 of 5 (100%) specimens infected with S aureus. It was concluded that the antibody-coated bacteria test does not provide accurate information to support the diagnosis of pyelonephritis in the dog.