Prenatal ultrasounds do not cause autism. Multiple large studies, including a meta-analysis covering more than 37,000 children with autism spectrum disorder, have found no link between ultrasound exposure during pregnancy and autism risk. This is one of the more thoroughly investigated questions in prenatal medicine, and the evidence consistently points in the same direction.
What the Research Shows
The most direct evidence comes from the Study to Explore Early Development (SEED), a large case-control study published in Paediatric and Perinatal Epidemiology. Researchers compared 1,524 singleton pregnancies, tracking both the number of ultrasounds each mother received and whether her child was later diagnosed with autism. The results were clear: increasing ultrasound counts showed no association with autism, whether the pregnancy had medical complications or not. The adjusted odds ratio was 1.01 in both groups, meaning the risk was statistically identical regardless of how many scans a mother had. Trimester-specific analyses using medical records showed no association in any individual trimester either.
An earlier finding did create some confusion. Children who were later diagnosed with autism tended to have had more prenatal ultrasounds than children without autism. But this turned out to be a classic case of correlation without causation. Those pregnancies also had more medical complications, such as gestational diabetes, preeclampsia, and fetal distress. The extra scans were a response to those complications, not the cause of anything that followed. Once researchers controlled for the medical reasons behind the scans, the apparent link disappeared entirely.
Why the Concern Exists
Ultrasound is a form of energy, not just a passive picture-taking tool. Sound waves bounce off tissue and return as an image, and in the process they can produce slight warming and mechanical vibration. The FDA has acknowledged that ultrasound “can’t be considered harmless, even at low levels,” which is a reasonable statement about any form of energy directed at a developing fetus. At higher power levels, ultrasound is used therapeutically to speed fracture healing and treat muscle injuries, precisely because of its tissue-warming effects.
Modern ultrasound machines display two safety metrics: a thermal index (indicating potential tissue warming) and a mechanical index (indicating the risk of mechanical effects like tiny bubble formation in tissue). The FDA sets a maximum mechanical index of 1.9 for diagnostic imaging, and research shows bubble formation doesn’t occur below 0.5. Standard prenatal imaging operates well within these safety margins. The energy levels used in a typical pregnancy scan are far below those used in therapeutic applications.
How Many Ultrasounds Are Recommended
The American College of Obstetricians and Gynecologists recommends at least one standard ultrasound during pregnancy, typically performed between 18 and 22 weeks. This anatomy scan checks for structural issues, confirms gestational age, and evaluates placental position. Some women also receive a first-trimester ultrasound, though this isn’t standard for low-risk pregnancies because it’s too early to see many fetal structures in detail.
High-risk pregnancies often require additional scans to monitor complications like restricted fetal growth or placental problems. The research from SEED specifically addressed this scenario and found no increased autism risk even with frequent scanning, which should be reassuring if your doctor recommends more monitoring than average.
Keepsake Ultrasounds Are a Different Story
The FDA has issued warnings about commercial “keepsake” ultrasound studios, which offer 3D and 4D video souvenirs of your baby. These businesses are not staffed by trained sonographers, and sessions can last up to an hour, far longer than a diagnostic scan. The concern isn’t autism specifically but rather unnecessary exposure to ultrasound energy with no medical benefit. The FDA previously shut down such studios and has considered further regulatory action.
The safety principle guiding medical ultrasound is sometimes called “prudent use.” A medically indicated scan should never be withheld or modified out of safety fears, because the diagnostic information it provides has real value. But repeated scanning purely for a nice portrait, or for sales demonstrations at trade shows, offers no medical benefit and qualifies as imprudent. If you want keepsake images, your standard anatomy scan typically provides opportunities for that without a separate commercial session.
What Actually Increases Autism Risk
A 2017 meta-analysis of 17 studies identified several factors associated with higher autism risk: advanced maternal age, gestational hypertension, preeclampsia, gestational diabetes, bleeding during pregnancy, breech presentation, and fetal distress. These are the same pregnancy complications that lead to additional ultrasound monitoring, which explains why early, less careful studies sometimes found a superficial correlation between more scans and autism diagnoses.
Autism has a strong genetic component, with heritability estimates consistently among the highest of any neurodevelopmental condition. Environmental risk factors tend to involve conditions that affect fetal brain development broadly, like severe maternal infection or significant nutritional deficiencies, not brief exposure to diagnostic sound waves. The energy delivered during a standard ultrasound exam raises local tissue temperature minimally and transiently, nowhere near the levels that would disrupt neural development.