Structural components operating at high temperature are often susceptible to the incubation and growth of cracks by creep. It is conventional to characterise this behaviour by subjecting laboratory specimens to a selection of constant loads and measuring the time taken for a pre-existing crack to extend a finite amount. In practice, structural components are subjected to external loading such that material at stress concentrations experiences boundary conditions that are a mixture of load and displacement control. Often these boundary conditions are associated with elastic follow-up. This paper describes a set of experiments undertaken to quantify the influence of elastic follow-up on creep crack incubation in a high carbon, high chromium, martensitic (P92) steel at 650°C. The experiments were performed using parallel bar test rigs designed in a previous research programme (Shirahatti et al [1,2]). The test rigs used modified compact tension C(T) specimens fitted adjacent to two parallel bars that introduced elastic follow-up as the C(T) specimen deformed.To complement the elastic follow-up experiments, a set of standard uniaxial round bar creep tests and constant load-controlled creep crack incubation tests were conducted. The uniaxial tests enabled observation of the elastic and creep response of the steel. These data were used as input to both analytical and finite element models of the elastic follow-up experiments.The crack incubation times obtained from the elastic follow-up experiments were found to be significantly longer than those for equivalent conventional constant load conditions. This was assumed to be a result of the load on the specimen relaxing during elastic follow-up. Finite element models of the experiments confirmed the effect of the load relaxation.