Concentration measurements, stable isotope signatures, and elemental ratios are extensively used as parameters in mixing models to identify the source material contributing to mixtures. Determining these contributions in complex mixtures necessitates the use of multiple tracers. Here, we show that this requires careful selection of the parameters used, and that the common direct use of isotopic signatures or elemental ratios in mixing models when combined with other tracers can lead to erroneous results. We further demonstrate that when the number of available tracers exceeds the number of source materials, finding the fractional contributions that best fit the data may benefit from normalization, and we present a simple measure indicating when the outcome of mixing model calculations is poorly constrained. Finally, as a practical example, we investigated the impact of parameter choices through a reanalysis of literature data on organic matter cycling along the Altamaha River estuary.