Risks of contamination of the major clinical anticoagulant heparin can arise from deliberate adulteration with unnatural or natural polysaccharides, including heparin from other animal sources, other natural products, or artifacts of manufacture, and these can escape detection by conventional means. Currently, there is no generally applicable, objective test recommended by regulators that can detect these in pharmaceutical heparin, and this continues to leave heparin exposed to contamination risks. Two-dimensional correlation spectroscopic-filtering with iterative random sampling (2D-COS-firs) is reported. It employs a difference covariance matrix with iterative random sampling, and is capable of revealing contamination in pharmaceutical heparin to a high level of sensitivity irrespective of the nature of those features. The technique is suitable to any situation in which a comparison of a single entity to a family of heterogeneous entities, particularly natural products and biosimilars, needs to be made, and will find application in pharmaceutical monitoring, manufacturing quality control, materials science, biotechnology, and metabolomic investigations.