This paper systematically investigated the interference of calcium and magnesium in protein measurement with a modified Lowry method first proposed by Frolund et al. (Appl Microbiol Biotechnol 43:755-761, 1995). This interference has in the past been largely ignored resulting in variable and unreliable results when applied to natural water matrices. We discovered significant formation of calcium and magnesium precipitates that lead to a decline in light absorbance at 750 nm during protein determination. Underestimation of protein concentration (sometimes even yielding negative concentrations) and low experiment reproducibility were demonstrated at high concentrations of divalent cations (e.g., [Ca2+] over 1 mmol L-1). To eliminate interference from calcium and magnesium, two pretreatment strategies were established based on cation exchange and dialysis. These pretreatments were convenient and were found to be highly effective in removing calcium and magnesium in protein samples. By using the modified Lowry method with these pretreatments, proteins in standard solutions and in wastewater samples were successfully quantified with good reliability and reproducibility. In addition, we demonstrated that simultaneous quantification of humic substances with the modified Lowry method was not affected by the two pretreatments. These approaches are expected to be applicable to protein and humic substance determination in different research fields, in cases where the modified Lowry method is sensitive to divalent cation concentrations.