Remains of animal fuel and driftwood fires are evident in Birnirk and Thule sites of northwestern Alaska (AD 11th-14th century). To better understand these fires, a robust experimental protocol was designed to study the effects of multi-fuel fires, in particular, the addition of fat to woody fuels. In Arctic regions, permafrost and climate conditions do not allow for the development of tree vegetation. Marine mammal oil and bones served as fuel substitutes, as did locally shrubby vegetation and driftwood accumulations. The excavation of numerous thick burnt areas in many Arctic sites confirms the use of multiple fuels including wood, animal fat, and bone in large quantities. These burnt areas correspond to a wide range of fire activities-cooking, smoking, firing ceramics, and others-but the actions and effects of each fuel are still poorly known. We describe conditions necessary to achieve a reproducible and statistically representative experimental fire sample. We compared fuel combinations of driftwood or non-drifted wood, animal fat, and caribou bones over 55 combustions. Experiments were conducted under controlled conditions in a laboratory in France and on the coast of northwestern Alaska. We found that a minimum of 30 test assays was needed to obtain statistically significant results but many research avenues can be obtained from smaller series. We obtained key figures and descriptive data on the impact of different animal fuels on fire temperature and duration, as well as on the firewood spectrum, with important implications for the representation of different woody fuels and the fragmentation patterns of charcoals. We report a relatively rapid rate of formation for blackened and crusted sediments when seal oil is burned along with driftwood. This means that thick accumulations of burnt material may not be a reliable signal of long-term occupations and that the relationship between the duration of site occupation and fuel management deserves further study.