A statistical study of thumb reconstruction after traumatic amputation, based on the records of 33 pollicisations and 21 microsurgical toe-to-hand transfers was made. A precise method of evaluation of results is presented. The types of lesions are divided into 4 groups; 1: amputation of the thumb without injuries of the other fingers; 2: amputation of the thumb associated with amputations or mutilation of other fingers; 3: metacarpal hand; 4: amputation of the thumb distal to the metacarpophalangeal joint. The aim of this study was to determine the respective indications of the two methods. It may be concluded that: in group 1, the results of pollicisation and of toe-to-hand transfers are similar. The former is superior as far as the discriminative sensory and fine motor results are concerned, while toe transfer reestablishes better strength. In group 2, the more the other fingers are mutilated, the more toe transfers should be preferred to pollicisation which weakens the performance of the long fingers, especially when they are injured. In group 3 (metacarpal hand), the transfer of one or more toes is a revolution. It is the only technique capable of returning function and, if technique is correct, a cosmetically satisfactory aspect as well. In distal amputations of the thumb (group 4), the only indications for pollicisation are cases of a proximally injured finger with a healthy distal segment, transferable on healthy pedicles, which is relatively rare. Partial toe-to-hand transfers, is a new solution to this challenge.