© 2019 Elsevier B.V.Estimating the abundance of arbuscular mycorrhizal fungi relies entirely on indirect methods, meaning all measures are associated with some variability. The most common methods use microscopic estimates of the relative proportion of root length colonized by fungal structures. These methods typically examine root subsamples. While such methods are inexpensive and relatively simple, significant variation within single root system means there is opportunity for sampling bias. We evaluated the two most common methods of percent root length colonization for AM fungi both as a subsample and for the entire root system of flax plants. We compared these measures to a novel technique that returns projected fungal surface area (fungal coverage), by using microphotography and imaging analysis. Both microscopic methods overestimated the colonization intensity compared to image analysis. Among the microscopic methods, the method which incorporated colonization intensity (Trouvelot) was significantly more similar to imaging method results, than the method that is based on the presence/absence of the fungus (McGonigle).