The technique of spectrophotometric determination of analginum, caffeine, and paracetamolum in medicinal preparations Solpadeine and Tetralginum with the use of three sets of analytical wavelengths (AWL) selected using a new criterion is developed. Components with small and large contents can be determined approximately with an identical error. Phenobarbital and codeine do not interfere with the determination of the basic components.The Vierordt method is used for spectrophotometric analysis of multicomponent mixtures with overlapping spectra [1]. The accuracy of this method, which is based on solving a system of linear equations, depends mainly on the correct choice of analytical wavelengths (AWL). The numerous criteria for selecting the AWL that exist today have one common feature. It is assumed that additivity in the studied (henceforth for simplicity, binary) system is ideally fulfilled, i.e., the optical density of the mixture (A X+Y ) is exactly equal to the sum of the optical densities of the components (A X + A Y ) measured separately over the whole wavelength range [1 -6]. Small deviations from additivity that are observed in practice (DA = A X+Y -A X -A Y ) are considered to be random errors in the measurements. In this instance, they can be avoided by repeating measurements and averaging the results.However, in our opinion, deviations from additivity caused by systematic factors can appear in actual analytical methods. For example, these factors could be systematic errors of the measurement apparatus and the unaccounted influence of other mixture components, e.g., in analyzing multicomponent drugs the deviations may be related to the effect of the filler or other active component present at very small amounts and to the formation of ionic associates between the determined compounds. Obviously, such effects should produce systematic analytical errors.According to the literature, the effect of nonadditivity for a single system (X + Y) may appear to different extents depending on the wavelength [7]. Then, those portions of the spectrum in which nonrandom deivations from additivity appear must be avoided if the optimal set of AWL is selected.Furthermore, such deviations may have small absolute values. Therefore, it may also be that they have practically no effect on the analytical results of the mixture by the Vierordt method. It is important only to know beforehand what values of DA cr produce errors in the analytical results that exceed (even for one component) a certain allowed level. The effect of systematic factors does not interfere with the analysis of the mixture by the Vierordt method if wavelengths at which deviations from optical density will be less than DA cr are selected as the AWL. Knowing the concentration of components and the absorption coefficients at the analytical wavelengths, DA cr can be established preliminarily through a computer experiment for each actual mixture.In Part 1 of this series, we proposed a new criterion for selecting the AWL that took into account the coefficient of o...