Biorelevant dissolution test methods for lipid formulations of RZ-50, an experimental Roche compound, were developed and compared with standard compendial methods in terms of their in vivo predictability. Release of RZ-50, a poorly soluble weakly acidic drug, from lipid suspensions filled in soft gelatin capsules was studied in compendial and biorelevant media using the USP Apparatus 2 (paddle method) and the USP Apparatus 3 (Bio-Dis method). Pharmacokinetic data were obtained in dogs after oral administration of a single 2.5 mg dose of RZ-50 soft gelatin capsules in the postprandial state. Level A IVIVC analysis and curve comparison of fraction drug dissolved vs. absorbed using the Weibull distribution were used to evaluate the in vitro methods in terms of their ability to fit the in vivo plasma profiles. Very low drug release was observed with the paddle method owing to poor dispersibility of the lipids in the dissolution media, whereas the Bio-Dis method hydrodynamics facilitated release of the drug by emulsifying the formulation in the medium. The best IVIVC was obtained using a dissolution medium representing fed gastric conditions in combination with the Bio-Dis method. Curve comparisons of the fraction drug absorbed and the fraction drug dissolved profiles based on Weibull distribution fits yielded similar results. The Bio-Dis/biorelevant in vitro method appears to be suitable for this type of lipid formulation.