Quantitative protein determinations in routine laboratories are today most often carried out using automated instruments. However, slight variations in the assay principle, in the programming of the instrument or in the reagents may lead to different results. This has led to the prerequisite of method optimization and standardization. The basic principles of turbidimetry and nephelometry are discussed. The different reading principles are illustrated and investigated. Various problems are identified and a suggestion is made for an integrated, fast and convenient test system for the determination of a number of different proteins on the same instrument. An optimized test system for turbidimetry and nephelometry should comprise high-quality antibodies, calibrators, controls, and buffers and a protocol with detailed parameter settings in order to program the instrument correctly. A good user program takes full advantage of the optimal reading principles for the different instruments. This implies--for all suitable instruments--sample preincubation followed by real sample blanking, which automatically corrects for initial turbidity in the sample. Likewise it is recommended to measure the reagent blank, which represents any turbidity caused by the antibody itself. By correcting all signals with these two blank values the best possible signal is obtained for the specific analyte. An optimized test system should preferably offer a wide measuring range combined with a wide security range, which for the user means few re-runs and maximum security against antigen excess. A non-linear calibration curve based on six standards is obtained using a suitable mathematical fitting model, which normally is part of the instrument software.