Abstract Economic constraints are such that it is unlikely a pure surfactant will be used for major enhanced oil recovery projects. However, it is possible to manufacture at competitive prices classes of syntheic and natural petroleum sulfonates that have fairly narrow molecular-weight distributions. Under some reservoir conditions, one of these narrow-distribution sulfonates may serve quite well as the basic component of a surfactant flood, however, in many instances a mixture of two or more of these may be required. Since evaluation of a significant subset of "all possible combinations" is a formidable undertaking screening techniques must be established that can reduce the number of laboratory core floods required. It is well known that interfacial tension plays a dominant role in surfactant flooding. It has recently been shown that minimal interfacial tensions occur at optimal salinity, Cphi, where the solubilization parameters VO/Vs and Vw/Vs are equal. Additionally, it has been shown that interracial tensions are inversely proportional to the magnitude of the solubilization parameters. This paper demonstrates that optimal salinity and solubilization parameters for any mixture of orthoxylene sulfonates can be estimated by summation of mole-fraction-weighted component properties. Those properties, which could not be properties. Those properties, which could not be measured directly, were obtained by least-squares regression on mixture data. Moreover, for surfactants of known carbon number distributions, equations that are linear in mole fractions of components and logarithmic in alkyl carbon number were found to be excellent estimators of both Cphi and solubilization parameters evaluated at Cphi. parameters evaluated at Cphi. Optimal salinity and associated solubilization parameters were measured using constant weight parameters were measured using constant weight fractions of alcohol cosolvents and mixtures of seven products with narrow molecular weight distributions. The average alkyl carbon number of these products varied from about 8 to 19. Alkyl chain lengths of individual surfactant chemical species ranged from 6 to 24 carbon atoms. Introduction Optimal salinity and the amounts of oil and water contained in a microemulsion have been shown to play important roles in obtaining low interfacial tensions and high oil recoveries. Since economics of enhanced oil recovery projects demand use of inexpensive surfactants, broad-distribution products likely will be chosen. Knowledge of how to estimate optimal salinity and oil-water contents of microemulsions prepared from such products would reduce time involved in laboratory screening procedures. This paper presents a method for procedures. This paper presents a method for obtaining such estimates that should prove useful for all types of surfactant mixtures that involve homologous series. The basic concept used is that a given property of a mixture of components (Yi) is related to the sum of products of mole fraction of components in the mixture (Xij) and the "mixing value" of the property in question for that component (Y'j). In property in question for that component (Y'j). In other words, (1) This approach is similar, for example, to the pseudocritical method used by Kay to calculate pseudocritical method used by Kay to calculate gas deviation factors at high pressures. The properties of interest in this paper are optimal properties of interest in this paper are optimal salinity and solubilization parameters, Vo/Vs, and Vw/Vs, at optimal salinity. Two separate approaches were developed that depended on the degree of detail of the available surfactant-composition data. In the first approach, only average molecular weights of several surfactant products were assumed known. Optimal salinity and products were assumed known. Optimal salinity and solubilization parameters could be measured for some, but not all, of the products. Regression on mixture data was used to estimate these quantities for the remainder of the products. Those properties, either measured experimentally or estimated from mixture data, are referred to as surfactant product contributions since they can be used as mixing values of the property in question in Eq. 1 or Eq. 2. SPEJ P. 193