Genetic influence shapes who develops posttraumatic stress disorder (PTSD) after traumatic events. However, the genetic variants identified for PTSD may in fact be associated with traumatic exposures (e.g., interpersonal violence), which appear heritable as well. Childhood cancer survivors (CCS) are at risk for PTSD, but genetic influences affecting cancer are unlikely to overlap with those affecting PTSD. This offers a unique opportunity to identify variants specific to PTSD risk. In a genome-wide association study (GWAS), 3984 5-year survivors of childhood cancer of European-ancestry from the Childhood Cancer Survivor Study (CCSS) were evaluated for discovery and 1467 survivors from the St. Jude Lifetime (SJLIFE) cohort for replication. Childhood cancer-related PTSD symptoms were assessed using the Posttraumatic Stress Diagnostic Scale in CCSS. GWAS was performed in CCSS using logistic regression and lead markers were replicated/meta-analyzed using SJLIFE. Cross-associations of identified loci were examined between CCS and the general population. PTSD criteria were met for 671 participants in CCSS and 161 in SJLIFE. Locus 10q26.3 was significantly associated with PTSD (rs34713356, functionally mapped to ECHS1, P = 1.36 × 10–8, OR 1.57), and was replicated in SJLIFE (P = 0.047, OR 1.37). Variants in locus 6q24.3-q25.1 reached marginal significance (rs9390543, SASH1, P = 3.56 × 10–6, OR 0.75) in CCSS and significance when meta-analyzing with SJLIFE (P = 2.02 × 10–8, OR 0.75). Both loci were exclusively associated with PTSD in CCS rather than PTSD/stress-related disorders in general population (P-for-heterogeneity < 5 × 10–6). Our CCS findings support the role of genetic variation in PTSD development and may provide implications for understanding PTSD heterogeneity.