Critical-size bone defects up to 25 cm can be treated successfully using the induced membrane technique established by Masquelet. To shorten this procedure, human acellular dermis (HAD) has had success in replacing this membrane in rat models. The aim of this study was to compare bone healing for smaller and larger defects using an induced membrane and HAD in a rat model. Using our established femoral defect model in rats, the animals were placed into four groups and defects of 5 mm or 10 mm size were set, either filling them with autologous spongiosa and surrounding the defect with HAD or waiting for the induced membrane to form around a cement spacer and filling this cavity in a second operation with a cancellous bone graft. Healing was assessed eight weeks after the operation using µ-CT, histological staining, and an assessment of the progress of bone formation using an established bone healing score. The α-smooth muscle actin used as a signal of blood vessel formation was stained and counted. The 5 mm defects showed significantly better bone union and a higher bone healing score than the 10 mm defects. HAD being used for the smaller defects resulted in a significantly higher bone healing score even than for the induced membrane and significantly higher blood vessel formation, corroborating the good results achieved by using HAD in previous studies. In comparison, same-sized groups showed significant differences in bone healing as well as blood vessel formation, suggesting that 5 mm defects are large enough to show different results in healing depending on treatment; therefore, 5 mm is a viable size for further studies on bone healing.