Purpose: To investigate the association between maternal smoking in pregnancy, early-life environment and childhood vision. Methods: Twin and triplet children enrolled in the Twins Eye Study in Tasmania underwent a comprehensive ophthalmic examination and their parents/guardians retrospectively answered a questionnaire regarding crawling, walking and other measures. A subset of these twins was also in the Tasmanian Infant Health Survey, which prospectively collected data on antenatal smoking, gestation, birth weight and other factors. Results: The mean age of the 346 individuals (172 multiple birth sets) at the time of examination was 9.25 ± 2.4 years. Mean unaided visual acuity was 0.0 (6/6). The mean spherical equivalent was +0.87D, and decreased with increasing child age (p <0.01). A prospective analysis, accounting for birth set clustering and relevant confounders, showed increasing levels of maternal smoking in the third trimester was associated with poor stereoacuity on the Titmus test (worse (>) than 100″, p = 0.05) and Lang test (p = 0.001) and also with the presence of esotropia (p = 0.02). These associations persisted after adjustment for infant postnatal smoke exposure at one month of age. Poor stereoacuity on Titmus stereo test circles was associated with late age of first crawling (RR = 1.23 (1.06, 1.42) p = 0.005 per month) and late age of first walking (RR 1.18 (1.05, 1.22) p = 0.001 per month). Conclusions: Antenatal smoking was independently associated with poor stereovision and the presence of esotropia. Poor stereoacuity may be associated with delayed age at first crawling or walking.