This work presents the potentiality of partial least squares discriminant analysis (PLS-DA) associated with mid-infrared spectroscopy to detect gasoline, residual automotive lubricant oil, soybean oil, and used frying oil, in biodiesel/diesel blend (B5). The samples of biodiesel/diesel blend unadulterated and adulterated were classified correctly in their respective groups; that is, the PLS-DA models showed 100% correct classification in samples of the test set with high levels of sensitivity and specificity to discriminate between adulterated and unadulterated samples. These results indicate that the methodology proposed is a viable alternative to detect these types of adulterants/contaminants in biodiesel/diesel blend (B5), commonly used in Brazil.