Extravagant claims have been made repeatedly in recent years that human spe
rm counts are falling and that global exposure to environmental estrogens a
re responsible. The basis for these two distinct claims is reviewed. The cl
aims of falling human sperm output, reviving an old debate, are prompted by
a paper by Carlsen et al. (1992). This meta-analysis, however, is marred b
y numerous flaws that invalidate its claims. Major defects include severe h
eterogeneity of component studies, rendering them unsuitable for aggregatio
n, and defective data analysis based on arithmetic mean rather than median,
which showed no significant changes over time. This debate is likely to re
main unresolved until valid, representative population-based studies of hum
an sperm output can be achieved. None have been reported, or seem feasible
in the near future, and so alternative strategies, based on surrogate varia
bles for human male fertility not requiring sperm counts, need to be develo
ped and validated. The plausible hypothesis that prenatal estrogen exposure
might influence development of the human testis through effects on Sertoli
cell replication and sperm carrying capacity has, however, been conclusive
ly refuted by studies of boys born to women exposed to high doses of oral d
iethylstilbestrol during pregnancy. Neither fertility nor sperm output were
adversely influenced by massive maternal estrogen exposure during pregnanc
y, although minor urogenital malformations did occur. The still wider claim
s of deteriorating male reproductive health, notably changes in prevalence
or incidence of hypospadias or cryptorchidism, also lack convincing populat
ion-based evidence, although cancer registry data indicate a gradual increa
se in testis cancer in some countries. In summary, the available evidence d
oes not support claims of falling sperm counts or any general deterioration
in male reproductive health. Population-based studies of valid surrogate v
ariables for male fertility not requiring semen analysis are needed. If pop
ulation-based evidence regarding male fertility or sperm output could be ge
nerated, it is highly unlikely that prenatal estrogen exposure could be a v
alid explanation of any deterioration as massive maternal exposure to oral
estrogen has negligible effects on male fertility or sperm output.