Analysis of U in the samples containing a significant proportion of ²³²U and high concentration of Th is of great concern. Transmutation of Th in the nuclear power reactor produces a notable quantity of ²³²U (half life 68.9 years) along with fissile isotope ²³³U. The decay series of ²³²U is initiated with ²²⁸Th (half life 1.9 year) and it is followed by several short lived α emitting progenies, ²²⁴Ra, ²²⁰Rn, ²¹⁶Po, ²¹²Bi and ²¹²Po. Even at the smallest contamination of Th in the sample, a very high pulse rate of α emission is obtained, which is to be counted for the radiometric determination of [U]. A commercially available anionic type  of  extractant  Alamines^®336 is  used  to obtain  the  selective  extraction of  U  from  other alpha  active elements and fission products present in the sample. Experimental conditions of liquid-liquid extraction (LLE) are optimized for obtaining maximum decontamination and recovery of U in the organic phase. The effect  of  some  interfering  ionic  impurities  in  the  sample  on  the  process  of  separation  is investigated. Depending  on  the  level  of  the  concentration  of  U  in  the  samples,  spectrophotometry  or  radiometry methods  are  adopted  for  its  determination  after  separation  by  LLE.  Under  optimized  experimental conditions,  i.e.  5.5 M  HCl  in  the  aqueous  phase  and  0.27 M  Alamins^®336  in  the  organic  phase,  the  re- covery of U is about 100%, the decontamination factor with respect to Th is  > 2000 and the extraction of fission products like ⁹⁰Sr, ¹⁴⁴Ce and ^134,137Cs is negligible. The detection limit for [U] using α radiometry is 10 mg/L, even in presence of > 100 g/L of Th in the sample. Accuracy and precision for the determination of U is also assessed. Reproducibility of results is within 5%. This method shows very good agreement with  the  results  obtained  by  mass  spectrometry.