1) Twenty-three cases of complete duplication of the urinary tract experienced at our institution from 1970 to 1976 were reviewed.<br>2) They were subdivided depending on whether the major lesion of urinary obstruction was at the ureterovesical junction or in the kidney. Twenty-two cases belong to the former and were grouped into the following three groups: Group I of 9 cases with ectopic ureterocele, Group II of 9 cases with ureteral reflux (reflux to lower ureter in 7 and to upper ureter in 2) and Group III of 4 cases with ectopic ureter with marked hydroureter without ectopic ureterocele, ureteral reflux or stricture of the ureteric orifice. One special case was designated as Group IV because it was a case of intrarenal obstruction of upper pelvioureteric junction by vascular commpression.<br>3) Retrospective observation was made from the operative and clinical findings regarding the type of vesical junction of the twin ureters. The twin ureters which shared the common hiatus of the vesical wall were arbitrarily designated as Type C (common). This type was comprised of in total 16 cases (each 7 from Groups I and II and each one from Group III and IV.). The twin ureters which independently had a completely separate hiatus were designated as Type S (separate). Here the orthotopic ureter was anchored to the lateral cornu of the trigone through the cranial vesical hiatus, while the twin mate ectopic ureter, being separated from the orthotopic ureter, took a completely independent extravesical course and was anchored at the bladder neck and urethra through the different hiatus. This type was found in 4 cases (each 2 from Groups I and II.). The twin ureters which simulate Type S, but differ from this in that the twin mate ectopic ureter, after separating from the orthotopic ureter, coursed in the tunnel between the two layers (outer detrusor longitudinal and inner deep trigonal) of bladder base before being finally anchored at the bladder neck through the separate hiatus, were designated as Type I (intermediate). Three cases of the Group III belonged to this type. To take a scrutinizing look at the type of hiatal condition in the duplex ureters appears to provide a better clue in understanding various pathophysiological conditions including the kidney, ureter and ureteral end. Because of not an infrequent instance of submucosal ureteral extension from the hiatus in ectopic ureters, hiatal position and size, rather than the position of the ureteral orifice per se as postulated by Mackie and Stephens (1975), appears to be correlated better with renal dysgenesis.<br>4) Full utilization of batteries of diagnostic modalities from nephrotomography, voiding cystourethrography and endoscopy along with a shrewd suspect of the duplex system was of utmost importance in the diagnosis. These are also helpful in conceiving preoperatively about the type of the hiatus, the knowledge of which is valuable in the management of the ureteral end at the time of operation.<br>5) Twenty-one of 23 cases required some forms of operative treatment to be relieved of various disabilities from the duplex anomalies. This implies that the complete duplex of the urinary tract is no longer the same clinical entity, hitherto conceived as common but clinically insignificant one. Regarding the operative management the significance of not leaving excess ureteral stump was discussed, and our method of excising the ectopic ureteral stump was also described in detail.<br>6) In refecance to the embryogenesis of ectopic ureterocele, ureteral reflux and ectopic ureter o the duplex system, it was proposed to conceive them as developmental phenomenon involving the terminal end of the ureters where ureteral buds, dually layered precursors of bladder base and primitive vesicourethral canal interact to form the ureteral end in a true sence from the hiatus through submucosal ureter to the ureteral orifice. In understanding the abnormalities of the ureterovesical junction this e