The study aimed to examine the relationship between secondhand smoke (SHS) exposure and severe headaches or migraine in never-smoking adults verified by serum cotinine. Current evidence about the association between self-reported SHS exposure and headaches or migraine is limited and contradictory. An important issue lies in the lack of actual SHS exposure assessment through biomarkers. We conducted a cross-sectional study on 4560 never-smoking adults from the National Health and Nutrition Examination Survey (NHANES), 1999-2004. The SHS exposure was evaluated by measuring serum cotinine concentrations. The information regarding severe headaches or migraine was based on self-reporting. The overall prevalence rate of severe headaches or migraine was 20% (919/4560). After adjusting for relevant covariates, we found that heavy SHS exposure (serum cotinine at 1 to 10 ng/mL) was positively associated with severe headaches or migraine (OR: 2.02, 95% CI [1.19, 3.43]); however, no significant association was found between low SHS exposure (serum cotinine at 0.05 to 0.99 ng/mL) and severe headaches or migraine (OR: 1.15, 95% CI [0.91, 1.47]). Restricted cubic spline analysis showed that the natural logarithm of serum cotinine had a linear relationship with severe headaches or migraine (p = 0.335 for nonlinearity). Stratified analysis indicated that individuals with a BMI of <25 (p < 0.001 for interaction) and sedentary activity (p = 0.016 for interaction) modified the relationship between SHS exposure and severe headaches and migraine. Even after altering the definition of SHS exposure, excluding drugs that might affect the metabolism of serum cotinine, and multiple imputation, our sensitivity analysis results remained stable. The study demonstrated that heavy SHS exposure (serum cotinine at 1 to 10 ng/mL) had a significant positive association with severe headaches or migraine in never-smoking adults. Prospective studies are necessary to verify this relationship in the future.