This study aims to assess the validity and reproducibility of a culture-specific semi-quantitative food frequency questionnaire (FFQ) for Lebanese adults. The 94-item FFQ captures intake of traditional Mediterranean dishes and Western food, reflective of current Lebanese nutrition transition. Among 107 participants (18–65 years), the FFQ was administered at baseline (FFQ-1) and one year thereafter (FFQ-2); 2–3 24-h recalls (24-HRs)/season were collected for a total of 8–12 over four seasons. A subset ( n = 67) provided a fasting blood sample in the fall. Spearman-correlation coefficients, Bland–Altman plots, joint-classification and (ICC) were calculated. Mean intakes from FFQ-2 were higher than from the total 24-HRs. Correlations for diet from FFQ-2 and 24-HRs ranged from 0.17 for α-carotene to 0.65 for energy. Joint classification in the same/adjacent quartile ranged from 74.8% to 95%. FFQ-2-plasma carotenoid correlations ranged from 0.18 for lutein/zeaxanthin to 0.59 for β-carotene. Intra-class correlations for FFQ-1 and FFQ-2 ranged from 0.36 for β-cryptoxanthin to 0.85 for energy. 24-HRs carotenoid intake varied by season; combining season-specific 24-HRs proximal to biospecimen collection to the FFQ-2 improved diet-biochemical correlations. By applying dietary data from two tools with biomarkers taking into consideration seasonal variation, we report a valid, reproducible Lebanese FFQ for use in diet-disease research.