Infants with a clinical diagnosis of anophthalmia or microphthalmia were identified from four health registers in Sweden, covering different parts of the period 1965-2001. During the observation period, the rate of anophthalmia decreased from the early 1970s from 0.4 to 0.2 per 10,000 births. The registered rate of microphthalmia increased markedly during the observation period to reach a maximum in 1987 of about 1.5 per 10,000. About 10% of the 432 identified children had a chromosome anomaly. There was no geographical variation in prevalence and infants born in urban or rural districts had, if anything, a lower risk than infants born in cities (0.93 and 1.13 per 10,000, respectively). Non-eye malformations were more common at anophthalmia (63%) than at microphthalmia (30%). Sex ratio was normal and no statistically significant variation between sub-groups (anophthalmia, microphthalmia, isolated, associated with non-eye malformations) could be demonstrated. There was a marked risk increase with maternal age but no certain parity effect, no effect of maternal education, but a possible association with subfertility. Maternal smoking in early pregnancy seemed to increase the risk for anophthalmia or microphthalmia in the absence of a coloboma.