The present study describes the use of unbiased stereological methods for the quantitative evaluation of the amount of regenerated bone. Using the principle of guided bone regeneration the amount of regenerated bone after placement of degradable or non-degradable membranes covering defects in rabbit calvaria was compared. Forty rabbits were divided into 5 groups. A titanium microplate was placed over the defect to prevent collapse of the membrane. The non-degradable expanded polytetrafluoroethylene membrane and the degradable Polyglactin 910 material were both placed unicortically and bicortically. Undecalcified sections were prepared for stereologic evaluation after an observation period of 8 weeks. Complete bone healing of the defects was not observed in any of the specimens. Unbiased stereologic estimates revealed 48% bone regeneration in defects covered by 2 ePTFE membranes, and 12% in defects covered by 2 Polyglactin 910 membranes. Defects covered by 1 ePTFE or 1 Polyglactin 910 membranes revealed 10% or 18% bone regeneration, respectively. The control group regenerated 14%. The major difference of the estimates was caused by real difference between specimens, i.e. biologic variation, whereas only minimal variance was added by the stereologic estimation procedure.