It is already observed by many authors that the preliminary cold work, such as twist, tension, rolling, extrusion etc. will effect tile creep character of metal and alloy both in its transient (β- flow) and permanent (k-flow) stages. (l)-(8) However, their results are not consistant. In this article, the results of 23 creep tests of heat resistant alloy 10T2 (10% Cr and 2% Si Steel) with preliminary cold twist are reported. The angle of cold twist ranges from 0° to 650°, which approaches that at rupture (680°). The creep tests were carried cot on the Louts Schopper Dst 5000 creep testing machine under 500℃ and 12 kg/mm2. The duration of each test was 50 hours. The result of tests shows that: 1. All the values of creep rate of permanent stage of creep with cold twist are larger than that without cold twist. The variation of creep rate with the degree of cold twist forms’ nearly a smooth curve, in which 120° and 400° are two maximum points. 2. The full investigation of the effect of preliminary cold twist on the transient stage of creep is so complicate that no general relation can be found. These results are not in consistancy with those above mentioned previous works. It is already observed by many authors that the preliminary cold work, such as twist, tension, rolling, extrusion etc. will effect tile creep character of metal and alloy both in its transient (β-flow) and permanent (k-flow) stages (l) - (8) However, their results are not consistant. In this article, the results of 23 creep tests of heat resistant alloy 10T2 (10% Cr and 2% Si Steel) with preliminary cold twist are reported. The angle of cold twist ranges from 0 ° to 650 °, which approaches that at rupture (680 °). The creep tests were carried cot on the Louts Schopper Dst 5000 creep testing machine under 500 ° C and 12 kg / mm2. The duration of each test was 50 hours. The result of tests that that: 1. All the values ​​of creep rate of permanent stage of creep with cold twist are larger than that without cold twist. The variation of creep rate with the degree of cold twist forms’ nearly a smooth curve, in which 120 ° and 400 ° are two maximum points. 2. The full investigation of the effe ct of preliminary cold twist on the transient stage of creep is so complicate that no general relation can be found. These results are not in consistancy with those above said previous works.