Wild populations of Croatian clary sage (Salvia sclarea L.) were examined for variability to determine cultivation suitability in Croatia and Israel for breeding purposes. Phenotypic variability (coefficient of variation; %) was recorded for inflorescence weight (39.6%), inflorescence yield (52.8%), and essential oil yield (67.6%) when grown in Croatia. Associations were identified between inflorescence yield and essential oil yield (r = 0.9; P < 0.0001), inflorescence weight and inflorescence yield (r = 0.8; P < 0.0001), and inflorescence weight and inflorescence length (r = 0.6; P = 0.0056), suggesting that populations with elongated inflorescence are indirectly associated with higher essential oil yield. In Israel, the populations reached full bloom between the end of May and early June, corresponding on average to 397.5 days post planting. Linalyl acetate, linalool, α–terpineol, sclareol, and geranyl acetate were the leading essential oil components in both Croatia and Israel. The principal compounds in the oil were linalyl acetate (48.5%) and linalool (17.7%), signifying that the Croatian populations were of the linalool chemotype. A two-way ANOVA indicated an interaction between growing location (Croatia vs. Israel) and population for linalool (P = 0.02), α–terpineol (P = 0.007), and linalyl acetate (P = 0.09); evidence of an environmental effect on essential oil composition. The variation observed suggested that the wild population of clary sage in Croatia had the genetic heterogeneity essential for breeding. Nevertheless, the differences in essential oil composition between Croatia and Israel suggest that breeding efforts should be separately focused for each agriculture production system.