This is a forty-year story of my study of the development of clinical laboratory tests by mathematical approaches. First, I developed a digital filter which could remove components other than those of the QRS complex from a recorded electrocardiogram. This method is called the Okada algorithm and has been cited by many papers. Next, I studied the application of information theory and developed a function to reduce the amount of clinical data by eliminating relatively insignificant data items. This function was derived to give the minimum information quantity for the relative significance. The computed values were defined as "weighted entropy". During the last two decades, I have occupied myself with studies of clinical biochemistry. By investigating interactions between lipoproteins and surfactants, divalent cations, sugars, and lectins, we developed a new assay protocol to chemically measure the LDL cholesterol level in serum that does not require immunoseparation or centrifugation. The accuracy and precision of our assay protocol fulfilled the criteria of the NCEP Lipid Standardization Panel. The experimental data showed that our assay protocol was superior to the Friedewald formula. Also, I developed a homogeneous assay protocol to directly measure triglyceride content in VLDL. Possible reagents and conditions for measuring VLDL-TG were comprehensively tested, and the "best" combination was determined. Our data showed that the homogeneous method could successfully measure triglyceride content in VLDL particles. I hope that my experience of these studies will inspire younger investigators in the field.