The proportion of pregnancies exposed to either second-generation antipsychotics (SGAs) or first-generation antipsychotics (FGAs) varies between 0.3%–2% of all pregnancies, but, until now, little is known about the potential neurobehavioral teratogenicity of antipsychotics. Assessing this safety facet is the aim of this article. PubMed, Scopus, and Google Scholar were searched for eligible articles. PubMed (1954 to May 2016) was searched using several medical subject headings, variously combined. PubMed search results were also limited using the search filter for human studies published in English. Scopus and Google Scholar searches were filtered for article title (antipsychotics/neuroleptics, pregnancy). After excluding duplicates, 9,250 articles were identified and 29 met the following inclusion criteria: only articles that provided original/primary data on neurodevelopmental outcome in human offspring older than 4 months of age, independently of the study design, were selected for review. Indeed, some relevant neurodevelopmental milestones are achieved at this time. Length of study and neurodevelopmental assessment methodology did not influence the study selection. Unfortunately, published data on neurodevelopmental teratogenicity of SGAs mainly derive from case reports and small case-series studies. Even findings emerging from case-control and prospective/retrospective studies are of limited clinical relevance because of their small sample sizes. Limited data are also available on FGAs. Hence, we have to conclude that the long-term neurodevelopmental outcomes for children exposed in utero remain unclear. Low to very low quality evidence of retrieved data makes impossible to confirm or exclude potential long-lasting untoward effects on infant neurocognitive development associate with antenatal exposure to either SGAs or FGAs.