The treatment of the burned hand has always been a subject of special interest. In order to obtain a better understanding of the parameters involved in the long-term functional outcome of hand burns a retrospective study was performed on 88 consecutive patients with hand burns (143 hands), treated according to a standardised protocol. Patients were followed for at least 12 months postburn. Hand function was assessed by the seven objective test criteria (7-OTC) described by Jebsen. Logistic regression analysis produced five parameters that were found to have a significant predictive value for long-term hand function. In order of predictive value, these were finger amputations, age on admission, impaired autograft take, the full-thickness hand burn surface area and the full-thickness total body burn surface area. By fitting these five variables into an equation, a probability model was obtained, which could be applied to estimate a prognosis concerning the final hand function of an individual patient with a hand burn.No relationship was found between the postburn day of operation and the long-term hand function. This finding is inconsistent with the current consensus that functional outcome is improved by early excision and grafting. In practice, it suggests that hand function is well preserved when burns of uncertain depth are treated expectantly, followed by selective debridement and grafting. Advantages include reduced blood loss, no loss of vital tissue and a reduction of the need for donor sites. Copyright (C) 1999 Elsevier Science Ltd and ISBI.