The great amount of clinical analysis carried out daily in the hospitals justifies the attempts to develop fast screening tools in order to avoid the confirmation analysis of samples far from pathological values. Infrared (IR) is a promising technique on the clinical field that has been used for the determination of clinical parameters in sera. However, serum and related fluids are complex matrices which produce overlapped spectra and chemometric treatment is a necessary and critical step in order to extract as much as possible information. This chapter focuses on the extraction of the information about the concentration levels of clinical compounds in serum samples through chemometric treatment of IR spectra. To do it we have made a revision of the regression algorithms used for predicting the concentration of clinical parameters and evaluated the diagnostic capability of attenuated total reflectance (ATR) Fourier transform (FT)-IR measurements of serum samples for the detection of samples with normal values of target analytes using discrimination methods, such as linear discriminant analysis and partial least squares discriminant analysis.