The effect of radiotherapy (RT) and chemotherapy (CT) on gonadal function was assessed in males treated for a childhood brain tumor not directly involving the hypothalamus/pituitary (HP) axis in a population-based study with a long follow-up time. All males <15 yr at the time of diagnosis (median: 9.0 yr, range: 0.8 to 14.9 yr) and diagnosed from January 1970 through February 1997 in the eastern part of Denmark and [gte]18 yr at the time of follow-up (median: 25.8 yr, range:18.5 to 39.3 yr) were included. Thirty males fulfilled the criteria. The median age at time of RT was 9.0 yr (range: 0.8 to 14.9 yr) and the median length of follow-up was 18 yr (range: 2.0 to 28.0 yr). The biological effective dose of RT was determined to the HP region and to the spine and expressed in gray because the biological effective dose gives a means of expressing the biological effect on normal tissue of different dosage schedules in a uniform way. Levels of serum FSH, luteinizing hormone (LH), sexual hormone-binding globulin, testosterone, and inhibin B were measured and compared with healthy age-matched male controls (n = 347), and the patients had a GnRH stimulation test performed with determination of peak FSH and LH. Patients treated with RT + CT (n = 13), compared with patients treated with RT only (n = 17), had significantly higher median peak FSH (8.33 vs. 3.79 IU/L, P = 0.03) and median peak LH (20.0 vs. 12.8 IU/L, P = 0.03), and significantly lower median inhibin B (86.0 vs. 270 pg/ml, P = 0.03), and median inhibin B/FSH ratio (12.8 vs. 107.9, P = 0.04), which indicates gonadal damage. Inhibin B and inhibin B/FSH ratio were also significantly lower in the RT + CT group, compared with controls (median: 86.0 vs. 215 pg/ml, P = 0.02), (median:12.8 vs. 67; P = 0.01), respectively. We found a significantly inverse correlation between basal FSH and inhibin B and FSH and total testicular volume (r(s) = -0.83; P < 0.0001), (r(s) = -0.67; P < 0.0001), respectively, and a significant correlation between inhibin B and total testicular volume (r(s) = 0.63; P < 0.0001). Stepwise backward multiple linear regression analysis showed the best-fit model to predict inhibin B levels included total testicular volume (P = 0.002) and CT (P = 0.09). Median basal LH in the RT-only group was significantly lower, compared with controls (3.44 vs. 2.45 IU/L; P = 0.0001) indicating secondary hypogonadism, and in both the RT + CT group and the RT-only group, levels of testosterone were significantly lower, compared with our reference population (12.8 vs. 21.9 nmol/L; P = 0.001, and 14.7 vs. 21.9 nmol/L; P = 0.0003), respectively. In conclusion these data suggest that cranial irradiation for a childhood brain tumor may affect the HP axis, and adjuvant CT can reduce inhibin B indicating primary gonadal damage. Thus, such patients may have normal or even low levels of FSH despite damage to the seminiferous epithelium, and because the fertility status by a semen analysis for psychological reasons can be difficult to obtain in this group of patients, we suggest inhibin B as the most useful direct serum marker of spermatogenesis in the follow-up of individuals who have received both cranial irradiation and gonadotoxic chemotherapy. However, because the number of patients with RT + CT and RT only are small, these data must be confirmed in further studies.