Permanganate oxidizable carbon (POXC) is increasingly used in soil health assessments as an indicator of labile or active soil organic carbon (SOC). The POXC method owes its popularity to its rapidity and low cost and its responsiveness to soil management. However, the method's sensitivity to variation in methodological parameters may compromise the comparability of POXC values across soils. Here, we measured the effects of soil mass and the method of sample homogenization (grind size) on POXC values for a set of 42 soil samples, representing diverse geographic areas and soil types of the USA. Total SOC contents ranged from 3.0 to 288.4 g kg−1. Ten treatments, combinations of five sample masses (0.25–5.0 g) and two grind sizes (<2 mm, <0.18 mm), were evaluated for POXC based on Weil et al. (2003). Results showed that POXC values decreased exponentially with increasing soil mass used, while analytical variability decreased as well. Decreasing grind size from <2 mm to <0.18 mm increased POXC values by 49% and decreased analytical variability. Strong correlations between POXC values obtained at two different soil masses (r = 0.91–0.97) or grind sizes (r = 0.96) indicate that general trends in POXC values across soils are maintained irrespective of soil mass or grind size, but the results are not interchangeable since POXC values differ considerably. Therefore methodological parameters, including soil mass and intensity of soil homogenization, should be strictly controlled to ensure comparability. However, standardizing soil mass presents a challenge for its use in national monitoring schemes as none of the soil masses yielded results within the method's quantitation limits for all soils included in our sample set. A more fundamental problem challenging the comparison and interpretation of POXC values relates to the use of a fixed soil mass for POXC analysis across soils with different SOC contents. Whereas the initial amount of MnO4− is fixed, the amount of SOC that takes part in the oxidation reaction is soil specific and not controlled for when using a fixed soil mass. This leads to variation in the [MnO4−] : SOC mass ratio during the reaction, which affects the size and the biochemical composition of the fraction of total SOC being oxidized. We conclude that the POXC method falls short as an analytical method for measuring a well-defined C pool. Performing POXC analyses based on a fixed SOC mass may improve comparability of POXC values across soils with contrasting SOM contents, but would compromise the main advantages of the POXC method in terms of simplicity, speed and low-cost.