Exposure to occupational radiation can lower the male sex ratio. However, specific radiation exposure to the testes has not been evaluated. This study aimed to examine the association between testicular radiation exposure and lower male sex ratio in children. A comprehensive questionnaire survey was administered to 62 full-time male doctors with children aged < 10 years at 5 hospitals. Based on the possibility of testicular radiation exposure 1 year before the child's birth, participants were assigned to 3 groups as follows: RT (orthopedic surgery), RNT (cardiology/neurosurgery), and N (others). Intergroup differences in the proportion of female children were ascertained, and the female sex ratio (number of female/total number) of each group was compared against the standard value of 0.486. Multivariate logistic regression analysis with a generalized estimating equation was used to model the effects on the probability of female birth while controlling for the correlation among the same fathers. The study population included 62 fathers and 109 children, 49 were female: 19/27, 11/30, and 19/52 in the RT, RNT, and N group, respectively; the RT group had the highest proportion of females (p = 0.009). The p values for comparisons with the standard sex ratio (0.486) were 0.02, 0.19, and 0.08 for the RT, RNT, and N groups, respectively. Based on the N group, the adjusted odds ratios for the child to be female were 4.40 (95% confidence interval 1.60-2.48) and 1.03 (0.40-2.61) for the RT and RNT groups, respectively. Our results imply an association between testicular radiation exposure and low male sex ratio of offspring. Confirmatory evidence is needed from larger studies which measure the pre-conceptional doses accumulated in various temporal periods, separating out spermatogonial and spermatid effects.