Excretory pyelography has been found to be a valuable procedure in infants and children with renal insufficiency. Renal failure interferes with the radiographic definition of the urinary tract, but the information gained from even slight excretion of contrast material may be important in diagnosis, particularly if one wishes to avoid anesthesia or unnecessary instrumentation. To evaluate this concept, the experience of the last decade in The Montreal Children's Hospital has been reviewed. LITERATURE REVIEW Mosenthal and Lewis (16) assessed adult renal insufficiency on the basis of the nonprotein nitrogen level (NPN): normal, 40 mg. per 100 c.c. or less; slight impairment 41–45; moderate impairment 46–65; marked impairment 66–90; and maximal impairment 91 or more. Experience suggests that these values apply equally to children over one year. The reviews of Barnett and Vesterdal (1), and McCance (15) point out that in patients below the age of one year, renal function differs from that of the adult, and its evaluation is difficult. The kidney is not fully developed, the glomerular filtration and tubular concentration are less than in later life, and there is less renal reserve. One wishes to evaluate glomerular filtration by the NPN value, but it is also affected by protein intake and variations in the rate of urine flow. In the first year of life, dehydration commonly causes accumulation of nitrogenous waste products in the absence of renal disease. Assessment of the impairment of roentgen visualization of the urinary tract in renal failure has long been of interest as a possible method of evaluating renal function (2, 7, 8, 9, 13, 19) but has not proved to be a sensitive test, nor have measurements performed upon the urine excreted during and following intravenous pyelography (14). Attempts have been made to determine by clinical or laboratory criteria when excretory pyelography is likely to be useful. Campbell (4) states that results are apt to be unsatisfactory in patients of all ages when the NPN is more than 50 mg. per 100 c.c. Caffey (3) states that satisfactory visualization cannot be expected with values of 60 mg. per 100 c.c. or over, although late films (two to twenty-four hours) may afford diagnostic information. Josephson (11) does not define a figure but he is not afraid of giving Diodrast to patients with considerably impaired renal function, while Williams (21) feels that the blood urea should not exceed 100 mg. per 100 c.c. The review of Pendergrass et al. (18) and the experimental work of Olsson (17) on the hazards of intravenous pyelography show that renal failure does increase the risks of the procedure, particularly if there is liver failure or severe generalized disease. Braasch and Emmett (2), in a review of adults with renal insufficiency, considered the quality of the pyelogram in a variety of disease states. They found good correlation of renal function and excretory pyelography in hydronephrosis, pyelonephritis, polycystic kidney, and solitary cyst. Fairly good correlation was present with renal tumors and prostatic obstruction, while renal stones and tuberculosis often correlated poorly. A similar systematic study has not been done in children. METHOD OF STUDY The biochemical records of The Montreal Children's Hospital from July 1, 1951, until July 1, 1961, were searched and all those showing NPN values over 45 mg. per 100 c.c. of blood were selected. The charts and x-ray films of all the children who had intravenous pyelograms were reviewed. In 89, of whom 54 were under one year of age, pyelography was done when the NPN value was raised. Table I summarizes this experience. The nature of the azotemia was classified as prerenal, renal, or postrenal on the basis of history, physical findings, laboratory tests, intravenous pyelography, and follow-up evaluation. Division into diagnostic groups with further separation of patients under one year of age and over one year of age was made. The table also includes the range of NPN values and the diagnostic quality of the pyelogram.