Obtaining accurate estimates of remanent creep life is of great importance to the power generating industry. The small punch creep test promises to be a useful way forward in this respect. However, a major concern with the test revolves around the ability to convert small punch test data into the required uniaxial equivalents. Experimental results within the literature have given contradictory results partly due to the large experimental scatter inherent within the test and so this article reports some results from a recently developed stochastic finite element model of the small creep punch test that provides guidance on this matter. The uniqueness of the model is based on its realistic creep deformations laws, including strain hardening, thermal softening and damage accumulation that enables it to produce life predictions for virgin material as well as for material with pre existing damage. It is shown that the model produces excellent life predictions for virgin 0.5Cr-0.5Mo-0.25V steel and for damaged 1.25Cr-1Mo steel over a wide range of test conditions. The model also predicts that the dependency of the time to failure on minimum displacement rates is such that small punch test data can be converted into uniaxial data using relatively simple analytical expressions.