INTRODUCTION
Skin grafts are frequently used to reconstruct defects after
extirpation of lesions on the face, and are classified as either a full-thickness skin graft (FTSG) or a split-thickness
skin graft (STSG), according to the amount of dermis involved.1
Compared  with  STSGs,  full  thickness  grafts  include  the  entire
dermis and are better suited for reconstruction of facial defects
because they provide closer color, texture, and thickness matches. 1-3 Traditionally, the size of the FTSG is based on a template
from the defect or slightly larger.1-2However, when examining
long-term results it is not uncommon to find unsightly puckering of the mature graft beyond the borders of the defect (Figure
1). We believe that this deformity represents a misconception of
basic reconstructive principles, mainly  failure to  recognize the
naturally occurring wound contraction during healing and incorrectly planning an oversized graft. This notion is easily demonstrated by placing excised full thickness skin back in place immediately after it has been harvested: the wound margins drift
apart while primary contraction causes recoiling and shrinkage
of the excised skin, leading to a considerable discrepancy in size
(Figures 2 and 3).  
METHODS
Data from twenty cases of full thickness skin excisions was
reviewed  and  analyzed.  All  excisions  were  carried  out  according  to  the  principles  of  relaxed  skin  tension  lines.4-5
The  long  and  short  axes  of  both  the  excised  skin  and  the
open  excision  site  were  measured  immediately  after  harvesting.  Donor  sites  for  FTSG  were  primarily  sutured
without complications.
RESULTS
Patients were aged 29 to 81 years. Excision sites were categorized into 3 groups according to their anatomic region: half of
the cases were located on the face (eg, post and pre-auricular)
and 25% were located on either the extremities (eg, arms &
thighs) or torso (eg, back & abdomen) (Table 1). The minimal
average  discrepancy  between  the  excision  site  and  the  excised  skin  was  16%,  and  increased  to  32%  when  examining
the short axis. This finding was true for all three groups. The
average discrepancies regarding the short axes were at least
double than that of the long axes. This trend was most prominent on the extremities (Chart 1). 
                     
						





