Humans have systematic and reliable color preferences. The dominant account of color preference is that individuals like some colors more than others due to the valence of objects that they associate with colors (Ecological Valence Theory). In support of this theory, Palmer and Schloss show that the average valence of objects associated with a color, when weighted (the WAVE), explains up to 80% of the variation in color preference for adults from the United States (US). Here we investigate whether Ecological Valence Theory can account for the color preferences of female and male adults from Saudi Arabia to test how well the theory generalizes across cultures and how well it accounts for sex differences in color preference. We also extend the investigation of EVT by investigating whether abstract concept associations as well as object associations can account for preference. Saudi adults' color preferences, color object and concept associations, and association valence ratings were collected, and the WAVE was computed and correlated with preference ratings. The WAVE accounted for no more than half of the variance in Saudi color preferences, although there was some degree of sex specificity in the relationship of the WAVE and color preference. Adding abstract concept associations did not account for more variance than object associations alone, but the number of abstract concept associations did account for a significant amount of the variance in color preference for females, but not males. The findings converge with other cross-cultural studies in suggesting that the success of EVT in accounting for color preference varies across cultures and indicates that additional factors other than color associations are likely also at play.