Background: N-3 fatty acids in blood are adequate biomarkers of intake compared to dietary questionnaires because they do not rely on memory or self-reported information, and are not subject to interviewer bias. However, obtaining blood samples is an invasive approach, particularly among children. Measuring fatty acids in saliva is a non-invasive alternative that has not been explored yet in detail in population studies. Objective: To evaluate if n-3 fatty acids measured in saliva would be adequate biomarkers of intake. Methods: We selected 60 families for a 4-week intervention trial among participants in the Bogotá School Children Cohort. They were randomly assigned to receive a month’s supply of either soybean oil (high in α-linolenic acid [ALA], n=30) or sunflower oil (low in ALA, n=30). We obtained saliva and finger-prick blood samples in the children at baseline and follow-up. Fatty acids were analyzed by gas liquid chromatography in saliva and blood. To measure short-term responsiveness, we compared means of ALA in saliva by intervention arm at follow-up. To assess validity, we compared n-3 fatty acids concentrations in saliva and blood at baseline with the use of Spearman correlation coefficients. Results: Mean (SD) age of study participants was 14.5 (1.5) years; 37% (22 of 60) were male. Treatment groups did not differ at baseline with regard to socio-demographic and anthropometric characteristics, or fatty acids levels. Children assigned to soybean oil had a non-statistically significantly higher mean (SD) ALA concentration in saliva at follow-up (0.62 [0.66]), compared to children assigned to sunflower oil (0.47 [0.44]; p=0.45). Follow-up concentrations of long-chain n-3 fatty acids (EPA and DHA) in saliva did not differ significantly between intervention arms. To assess validity, we compared saliva and blood fatty acids concentrations at baseline (n=60) and found significant correlations for ALA (r=0.44, p=0.0005), EPA (r=0.27, p=0.04), and DHA (r=0.34, p=0.008). Conclusions: N-3 fatty acids measured in saliva do not reflect accurately short-term changes in ALA intake. However, moderate correlations between n-3 fatty acids in saliva and blood suggest that measuring fatty acids in saliva may be a valid proxy for intake. It is possible that fatty acids in saliva may have a slower turnover than those in blood and could be better biomarkers for medium or long-term intake of n-3 fatty acids.