Static formation temperature (SFT) can be estimated from temperatures measured during wire-line logging (T_m). A large number of correction models for obtaining SFT from T_m have been suggested. Several studies have shown that SFTs yielded by such models are off by an average of 6°C–10°C (43°F–50°F) at burial depths of 1.5–3.5 km (0.9–2.2 mi) and thus have the potential to cause serious issues in thermal and hydrocarbon generation models. This paper explores the causes for erroneous SFT predictions generated from T_m measurements and identifies factors that should be addressed to generate a globally applicable correction model. We also present an improved empirical correction model for T_m data from eight oil and gas fields, located on the Norwegian continental shelf. The new empirical model was designed to give correct average SFT predictions and is applicable to single T_m measurements. It has been validated against temperatures recorded during drill-stem testing, which closely represent local SFTs. The expression yields improved results compared with other correction models applied to the data set. However, the average error in computed SFT values varies by up to 10°C (18°F) between the investigated hydrocarbon fields. We conclude that these variations result from differences in operational practices such as fluid circulation and drilling velocities. Therefore, current empirical and physical models for SFT prediction from T_m require local calibration. It is also suggested that more accurate compilations and analyses of operational data could lead to improved and more globally applicable models.