In vitro tests for controlled release PLGA microspheres in their current state often do not accurately predict in vivo performance of these products during formulation development. Here, we introduce a new mechanistic and multi-phase approach to more clearly understand in vitro-in vivo relationships, and describe the first “in vitro phase” with the model drug, triamcinolone acetonide (Tr-A). Two microsphere formulations encapsulating Tr-A were prepared from PLGAs of different molecular weights and end-capping (18 kDa acid-capped and 54 kDa ester-capped). In vitro release kinetics and the evidence for controlling mechanisms (i.e., erosion, diffusion, and water-mediated processes) were studied in four release media: PBST pH 7.4 (standard condition), PBST pH 6.5, PBS + 1.0% triethyl citrate (TC), and HBST pH 7.4. The release mechanism in PBST was primarily polymer erosion-controlled as indicated by the similarity of release and mass loss kinetics. Release from the low MW PLGA was accelerated at low pH due to increased rate of hydrolysis and in the presence of the plasticizer TC due to slightly increased hydrolysis and much higher diffusion in the polymer matrix. TC also increased release from the high MW PLGA due to increased hydrolysis, erosion, and diffusion. This work demonstrates how in vitro conditions can be manipulated to change not only rates of drug release from PLGA microspheres but also the mechanism(s) by which release occurs. Follow-on studies in the next phases of this approach will utilize these results to compare the mechanistic data of the Tr-A/PLGA microsphere formulations developed here after recovery of microspheres in vivo. This new approach based on measuring mechanistic indicators of release in vitro and in vivo has the potential to design better, more predictive in vitro release tests for these formulations and potentially lead to mechanism-based in vitro-in vivo correlations.