This study aims to assess utility and limitations of grapefruit juice (GFJ) interaction studies as alternative in vivo approach to estimate intestinal availability (F(G)) in comparison to the predominantly used i.v./oral method. The F(G) estimates were obtained from the ratio of AUC in the control and the GFJ group reported previously. Due to large variability in the study design, the following inclusion criteria were applied for the selection of clinical studies: no change in elimination half-life in the presence of GFJ, administration of GFJ with or up to 4 h before drug intake, and a reported significant increase in AUC in the presence of GFJ. Weighted mean F(G) values were compared to estimates from i.v./oral data. Additionally, inter-study and inter-individual variation of GFJ F(G) estimates was assessed by meta-analysis for drugs with multiple studies reported. F(G) values ranged from 0.07 to 0.92 for lovastatin and quinidine, respectively. Overall, the inter-individual variation in GFJ F(G) estimates (16-54%) was higher than the inter-study (5.7-39%) with the exception of nisoldipine and simvastatin where inter-study variations of 53-88% were observed. Weighted average GFJ F(G) estimates were comparable to i.v./oral, supporting the application of this approach as an alternative to i.v./oral data for predominantly metabolised drugs (r(2) = 0.65; n=10). In contrast, this approach is of limited use for drugs whose disposition is co-dependent on efflux/uptake transporters and metabolic enzymes. An area of high intestinal extraction (F(G) < or = 0.25) is identified as problematic, as availability of conclusive data is limited in this area.