Previous evaluations of the Circumplex Model's curvilinear hypothesis using FACES instruments have yielded conflicting results. A review of the different research procedures and samples used in those investigations revealed that none of the studies had samples large and/or heterogenous enough to test the curvilinear hypothesis adequately. The present study evaluates the curvilinear hypothesis of family functioning and the concurrent validity of FACES III with a sample of optimal size (N = 2,440 families) and diversity. The lack of support for the curvilinear hypothesis in this “greenhouse” sample is explained by different findings for the two FACES III subscales. There was no relationship between the study's measures of well-being and the adaptability subscale and a linear relationship between these measures and the cohesion subscale. Implications of these findings for the continuing use of the FACES III and for the Circumplex Model of Marital and Family Systems are discussed.