In this opinion piece, we address the limitations of the two most common clinical endpoints in kidney transplantation trials (acute rejection and renal function) and attempt to offer a reasonable framework by which to find true and reliable early endpoints that reflect long-term outcomes. Other potential endpoints tested in recent years, including the use of genomic and proteomic markers are still in development. Until other reliable endpoints are established, it is important to understand what can be inferred from ongoing studies that utilize these endpoints and what further information we need to derive 'true' surrogate endpoints. We consider evaluation of current markers using the 'Prentice criteria', which bases assessment of endpoints as true surrogates on four primary rules. Based on our assessment, progress in understanding the safety and efficacy of new therapies and interventions in kidney transplantation will remain limited with current makers. Prospectively, we advocate: (i) significant caution in extrapolating long-term outcomes from currently utilized clinical markers, (ii) use of traditional hard endpoints whenever feasible and (iii) dedication of efforts for more data collection on specific disease entities and greater diligence in determining the onset of deleterious processes.