Bancroft's rule of thumb that a surfactant which preferentially partitions into water favors the formation of oil-in-water (O/W) emulsions and that a surfactant which preferentially partitions into oil favors the formation of water-in-oil (W/O) emulsions has been known for some 80 years. Although by judicious control of the emulsification process one can generate exceptions, the rule of thumb often works well. It was not, however, until Griffin's famous paper of 1949 that an attempt was made to quantify the emulsifying tendency of a surfactant. Griffin observed that certain combinations of the surfactant hydrocarbon and ethylene oxide chain lengths favored O/W or W/O emulsions. In the paper he noted that the emulsion chemist will choose a water-soluble surfactant to make an O/W emulsion and an oil-soluble one to make a W/O emulsion. To try to put this principle of hydrophile—lipophile balance (HLB) on a quantitative footing, he introduced the concept of an HLB number. According to Griffin, the HLB is the balance of the size and strength of the hydrophilic and lipophilic moieties of a surfactant molecule. The HLB number was based on the molecular groups composing the surfactant. According to this concept, the optimal surfactant for a desired emulsion of a given oil can be chosen by a simple calculation from a look-up table. A problem with the HLB concept was that the HLB numbers were based on ambient temperature data. A given ethoxylated hydrocarbon can form in O/W emulsions at room temperature while it gives W/O emulsions at higher temperatures. Thus the HLB cannot be related solely to the molecular groups of the surfactant. Temperature and interactions with the aqueous and oil phases have to be incorporated into the HLB number if it is to be a predictive tool. There is a strong correlation between whether an O/W emulsion or a W/O emulsion is formed and whether the surfactant tends to form an aqueous micellar phase or an oil-rich inverted micellar phase. Thus, as preferred by Becher, relating the HLB number to the free energies of micellization in water and oil phases ought to result in a more predictive tool. Shinoda and co-workers assert that it is preferable to introduce an HLB temperature based on the phase inversion temperature (PIT). This temperature is a property of the surfactant—oil—water phase diagram. For non-ionics, the PIT is the temperature below which a surfactant partitions preferentially into the water phase as oil-swollen micelles and above which it partitions preferentially into the oil phase as water-swollen inverted micelles. The advantage of basing the HLB on the PIT is that the surfactant—oil—water interactions are automatically incorporated into the HLB criterion. Similarly, the effect of cosurfactants, frequently needed for stable emulsions, is accounted for as a shift in the PIT. Finally, basing the HLB temperature on the PIT admits the strong correlation between emulsion type and ternary or pseudoternary phase behavior. We explore here the connection between the generic ternary phase behavior and emulsification tendencies and arrive at a generalization of Shinoda's PIT criterion. The trend from preferential partitioning of surfactant into the aqueous phase to a balanced or optimal microemulsion (equal uptake of oil and water) to preferential partitioning into the oil phase can be accomplished by a large number of variables (e.g. the field variables temperature, salinity and alcohol activity, and the extended field variables such as the average carbon number of the oil, the average carbon number of the hydrophobic moiety of the surfactant, the average number of ethylene oxide groups of the hydrophilic moiety of the surfactant, etc.) for the various types of surfactant. The field variable at which the optimal microemulsion occurs provides an optimal point (OP) that can serve as the HLB for a given system. If the field variable is set on one side (above or below depending on the variable) of the OP for a system, the surfactant serves as a hydrophilic agent (favors O/W emulsions) and if it is set on the other side of the OP, the surfactant serves as a hydrophobic agent (favors W/O emulsions). Since an optimal point can be achieved by variation of many parameters and field variables, the optimal microemulsion, and thus the HLB, is defined by a hypersurface in the space spanned by these variables. We also discuss the relationship between generic phase behavior of surfactant, oil and water systems and the curvature properties of surfactant interfacial films in association colloids.