Designing a fuel-sampling program that accurately and efficiently assesses fuel load at relevant spatial scales requires knowledge of each sample method's strengths and weaknesses.We obtained loading values for six fuel components using five fuel load sampling techniques at five locations in western Montana, USA. The techniques included fixed-area plots, planar intersect, photoloads, a photoload macroplot, and a photo series. For each of the six fuels, we compared (1) the relative differences in load values among techniques and (2) the differences in load between each method and a reference sample.Totals from each method were rated for how much they deviated from totals for the reference in each fuel category. The planar-intersect method, which used 2.50 km of transects, was rated best overall for assessing the six fuels. Bootstrapping showed that at least 1.50 km of transect were needed to obtain estimates that approximate the reference sample. A newly developed photoload method, which compared fuel conditions on the forest floor with sets of pictures calibrated for load by fuel type, compared well with the reference and planar intersect. The commonly used photo series consistently produced higher mean load estimates than any other method for total fine woody debris (0.05-0.20 kgm-2) and logs (0.50-1.25 kgm-2).