Although extensively studied (and of large importance for food processing) there is still no general consensus on the time-dependent convective heat transfer coefficient (h) in immersion frying - reported values differ by orders of magnitude (4-7000 W m(-2) K-1). Moreover, none of the many methods suggested have become a generally agreed upon standard. This study uses the general uncertainty management (GUM) framework to analyze the standard uncertainties in four commonly used methods to measure h in immersion frying. It is concluded that the uncertainties (when including all sources of uncertainty in a systematic manner using the GUM) are often substantially larger than previously reported (up to 100% relative uncertainty), i.e., some methods used in previous studies are very sensitive to measurement uncertainties. This can help explain the large variation in previously reported h-values. Comparisons between methods suggest that evaporation rate based methods (with numerical differentiation) provide the highest reliability.