AimsThis study aimed to investigate the association between serum levels of common and uncommon unsaturated fatty acids and prediabetes risk. MethodsData were collected from the National Health and Nutrition Examination Survey for 2003–2004 and 2011–2012. Weighted proportional and multivariate logistic regression analyses were performed to assess the association of serum PUFAs and MUFAs with prediabetes risk after adjusting for potential confounders. ResultsA total of 3575 individuals were enrolled in this study. Serum levels of PUFAs EPA (20:5 n3) and GLA (18:3 n6) were associated with increased prediabetes risk (EPA (20:5 n3): OR = 1.878, 95% CI: 1.177–2.996, Ptrend = 0.002; GLA (18:3 n6): 1.702, 95% CI: 1.140–2.541, Ptrend = 0.016). The MUFAs PA (16:1 n7) and EA (20:1 n9) were associated with the risk of prediabetes (OR in quintile5: PA (16:1 n7): 1.780, 95% CI: 1.056–3.001, Ptrend = 0.003; EA (20:1 n9): 0.587, 95% CI: 0.347–0.994, Ptrend = 0.010). Moreover, nonlinear analysis revealed that serum levels of EPA (20:5 n3) and EA (20:1 n-9) were nonlinearly associated with prediabetes risk. ConclusionSome serum n-3 PUFAs are positively associated with prediabetes, several serum n-6 PUFAs are inversely associated with prediabetes. Regulating individual serum USFA levels may help prevent prediabetes, thereby providing evidence for clinical and nutritional practices.