Objective. To present epidemiologic data on the relatively rare malfor
mation choanal atresia, based on a large collection of material and wi
th special stress on the significance of the so-called CHARGE (colobom
a, heart defect, choanal atresia, retarded growth and development, gen
ital anomaly, and ear defect with deafness) association. Methods. Data
from three large registries of congenital malformations were used. Ba
sed on more than 5 million births, 444 infants with choanal atresia we
re identified. Results. The average rate of choanal atresia is 0.82 pe
r 10 000 and varies among programs. There is no statistically signific
ant difference between races in rates, even though white infants have
a higher rate than those of other races. The higher rate found in the
California program is mainly attributable to unilateral, isolated case
s. Unilateral atresia occurs equally often on the right and left. Amon
g all cases of choanal atresia, the sex distribution is normal, a slig
htly increased risk at twinning exists, and no effect of maternal age
or parity is seen. Chromosome anomalies are found in 6% of infants wit
h choanal atresia, and 21 infants (5%) have monogenic syndromes or con
ditions. An analysis of associated malformations (present in 47% of th
e infants without chromosome anomalies) indicated that although a weak
nonrandom association can be demonstrated between the malformations e
ntering the so-called CHARGE complex, only a small proportion of infan
ts with choanal atresia and other components of that condition probabl
y represent this entity. The term CHARGE association seems to be overu
sed in clinical practice. Conclusion. To be meaningful, the term CHARG
E should be restricted to infants with multiple malformations and choa
nal atresia and/or coloboma combined with other cardinal malformations
(heart, ear, and genital) and with a total of at least three cardinal
malformations. Growth retardation should not be used in the definitio
n.