It is possible for hair to be externally contaminated by drugs like cannabis or cocaine, which are smoked or snorted. Three steps are commonly employed to minimize the chance of external contamination causing misinterpretation of the results of a hair test. The first consists of decontamination of hair samples by washing the hair before analysis, the second is the use of cut-off levels, and the third is the detection of both the parent drugs and appropriate levels of their metabolite(s) in the hair sample. We propose an additional step for the assessment of drug use using hair samples combined with decontamination data. Hair samples from 186 drug users were analyzed along with their wash residues by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The results of the hair analysis of the 140 samples for cocaine showed that 85.5% (N=89) of the samples passed 'cocaine use' criteria for metabolites ratios and 12.5% (N=13) for wash residue criteria (<10% of cocaine in the wash residue) leading to conclusive interpretation. Only two cases (1.9%) had an uncertain conclusion of drug consumption because cocaine levels in the wash residue were >10% of the levels in the hair. The results of the cannabis set of samples (N=46) were not as clear-cut, as a comparatively large number of samples (15.2%) had relatively high levels of THC in the wash residues. To use this approach, it is important that laboratories testing drugs in hair samples can demonstrate that the method utilized does not generate significant levels of the cocaine metabolites.