Children with an uncomplicated femoral fracture, treated with superimposition of fragments and intentional shortening, usually develop overgrowth of the fractured femur and the ipsilateral tibia which may compensate for the initial shortening and enable the limb in question to reach a length similar to that on the normal side. The overgrowth is evaluated clinically and by scanography. The increased metabolic activity of the growth plates that support this overgrowth has not been documented by any laboratory method. In order to evaluate the metabolic activity of the growth plates, 18 patients (11 males, seven females; mean age 6.1 years) with fractures of the femur were studied at three different time intervals (2-5 months, 6-12 months and 18-24 months). Three-phase bone scintigraphy was performed in all patients. Ten children (five males, five females; mean age 7.5 years) who had had bone imaging for other reasons were used as the control group. Visual analysis of the flow and equilibrium phases was performed for the distal femoral and proximal tibial growth plates. Visual and semi-quantitative analyses of the delayed images were performed for the distal femoral and proximal and distal tibial growth plates. Semi-quantitative analyses yielded the following activity ratios: (a) the distal femoral growth plate of the fractured femur to the contralateral one (FR); (b) the proximal growth plate of the tibia on the side of the fractured femur to the contralateral one (TpR); (c) the distal growth plate of the tibia on the side of the fractured femur to the contralateral one (TdR); and (d) in the control group, the distal growth plates of both femora (FCG) and the proximal (TCGp) and distal (TCGd) growth plates of the tibiae. Visual analysis of the blood flow, equilibrium and delayed images showed increased activity in the distal femoral growth plates during the first and second time intervals, but not during the third. No significant activity changes were found in the proximal and distal tibial growth plates during any of the phases analysed. The mean and standard deviation for FR in the three time intervals were: FRI=1.22+/-0.27, FRII=1.17+/-0.16 and FRIII=1.09+/-0.20. FR values were significantly higher than in the control group (FCG=0.99+/-0.03) (P=0.033). The mean and standard deviation for TpR in the three time intervals were: TpRI=1.08+/-0.18, TpRII=0.94+/-0.09 and TpRIII=0.96+/-0.20. TpR values were not significantly different from those in the control group (TCGp=1.00+/-0.05). However, TpRI was significantly higher than TpRII (P=0.043). The mean and standard deviation for TdR in the three time intervals were: TdRI=1.10+/-0.41, TdRII=1.05+/-0.15 and TdRIII=1.13+/-0.36. TdR values were not significantly higher than in the control group (TCGd=1.00+/-0.04) (P=0.777). These results support the concept that three-phase bone imaging is able to quantify and determine that activation occurs in the distal femoral and proximal tibial growth plates of fractured femora. This phenomenon may explain the overgrowth observed in this injured bone structure.