This paper assesses the relative accuracy of a number of techniques that are capable of predicting a wide range of creep properties. The techniques studied in this paper include the 4-Theta, 6-theta, CRISPEN, Omega and CDM methodologies. The parameters of these models were estimated from short-term creep property data on 1CrMoV steel and these estimated models were then used to predict the known longer-term creep properties of this steel. It was found that the CDM approach yielded predicted times to failure that were wholly inadequate. The 6-theta approach was best for predicting long term times to failure. The best minimum creep rate predictions came from using either the 4-Theta, or the CRISPEN or the Omega technique. Finally, times to small strains were best predicted using the 6-theta and CRISPEN techniques. (C) 2004 Kluwer Academic Publishers.