Thirteen institutions have received grants from the Japanese Ministry of Education, for comprehensive scientific research into vesico-ureternal reflux (VUR). On this research grant, we have made inquiries into 494 patients with non-obstructive VUR.<br>A computerized review of these cases was as follows:<br>A. Analysis of 494 cases (a real number).<br>1. There were 130 males and 364 females; the male-to-female ratio was 1:2.8.<br>2. The incidence of previous urinary tract infection (UTI) was 76 per cent in males and 88 per cent in females. This difference was caused by adult females under full, sexual life.<br>3. We grouped the VUR into six grades. Low grade reflux (IA-IIB) recovered in approximately 70 per cent and reflux less than III grade comprised more than 90 per cent. It became clear that patients with primary VUR mostly had low grade reflux. Patients with bilateral reflux comprised about 45 per cent of this series. High grade bilateral reflux tended to occur among young male patients.<br>4. The coefficient of correlation between right and left reflux was r_s=0.63. This indicated that primary VUR has a high probability of developing bilateral reflux.<br>5. As far as the coefficient of correlation was concerned, there were no relationship between the grade of reflux and the present illness.<br>6. The coefficient of correlation between grade of reflux and renal function using intravenous pyelography (IVP) was very low (r_s=0.258), whereas that between grade of reflux and morphological change of the pyelogram was very high (r_s=0.662). Therefore, high grade VUR was associated with morphological change rather than renal function.<br>B. Analysis of 536 cases (a total number). 42 of this series (494 cases) were included in both groups (group A and B) because surgical repairs were performed after long-term (over 6 months) chemotherapeutic treatment, so that the series amounted to 536 cases.<br>1. We divided 536 cases into four groups; i. e., group under chemotherapy (group A) 297 cases (55.4%), surgically treated group (group B) 194 cases (36.2%), non-treated group (group C) 39 cases (9.3%) and a fourth group (group D) 6 cases (1.1%).<br>2. A comparative study of the characteristics of A, B and C groups was as follows: Group A consisted of relatively young girls with low grade reflux. Group B was made up of a large number of cases with high grade reflux and recurrent UTI. Group C was composed of cases with low grade reflux and few episodes of infection.<br>3. Surgery for reflux consisted of combined Politano-Leadbetter's method 66 cases (23.1%), original Politano-Leadbetter's method 59 cases (20.6%), Lich-Gregoir's method 58 cases (20.2%) and Paquin's method with cuff 49 cases (17.1%).<br>C. Analysis of the results in each group.<br>1. Cures of reflux after each type of treatment were compared. Group B had significantly good results, especially grade III and IV groups. This rate differed from other groups to a statistically significant degree (p<0.001). Overall cure rate in group C was 70 per cent, which was better than in group A.<br>2. Concerning fever attacks, group B had the highest improvement rate, especially the cases with reflux greater than grade II. This was almost the same as the cure rate of reflux.<br>3. Concerning renal function, IVP results did not show any statistically significant difference among the three groups. However, the changes in pyelographic forms of group B, especially in the grade III, differed significantlly from other groups.<br>4. In group A, analysis of reflux-cured cases showed the following: These cases almost all have less than grade II reflux. All reflux-cured cases over grade III reflux were younger than ten years old.<br>5. The cure rate by original Politano-Leadbetter's method was 100 per cent, Paquin's method 94.6 per cent, combined Politano-Leadbetter's method 90.9 per cent and Lich-Gregoir's method 77.8 per cent. Regarding unsuccessful cases of surgery