Novel nonlinear multi-photon laser spectroscopic methods are presented as highly sensitive absorption-based detection method with a wide range of biomedical applications. Laser wave-mixing methods offer inherent advantages including excellent sensitivity, small sample requirements, short optical path length, high spatial resolution and excellent standoff detection capability. In a typical wave-mixing setup, the signal is generated when the two input beams intersect inside labeled or native label-free analytes. The signal is a coherent laser-like beam and can be collected with virtually 100% efficiency and minimal background noise. The signal has a quadratic dependence on analyte concentration, and hence, it is inherently suitable as a chemical sensor. For biomedical applications, the sensitivity levels are ideal for the detection of specific biomarkers such as those associated with multiple sclerosis (MS). The symptoms of MS are caused mainly by destruction of myelin in the central nervous system. Due to its similarity with many other neurological disorders, MS is currently diagnosed based on symptoms and confirmed by MRI images of the brain showing lesions. Sensitive chemical-based detection methods are needed in order to detect and diagnose MS before lesions grow to the size detectable by MRI. Laser wave mixing is reported as a reliable chemical-based detection system for possible early diagnosis of multiple sclerosis. For security applications, we demonstrate sensitive wave-mixing detection of triacetone triperoxide and ammonium nitrate explosives using capillary- and surface-based wave-mixing detection modes.