Prenatal ultrasounds are considered safe when performed by trained healthcare providers for medical purposes. Unlike X-rays or CT scans, ultrasound uses no ionizing radiation. It works by sending mechanical sound waves through tissue, which bounce back to create an image. Decades of use and multiple large studies have found no evidence of harm to mothers or babies when ultrasound is used according to standard medical guidelines.
How Ultrasound Differs From Radiation
The safety of ultrasound starts with its basic physics. X-rays and CT scans use ionizing radiation, which carries enough energy to break chemical bonds in DNA. Ultrasound does nothing of the sort. It produces a series of pressure waves that travel through tissue, alternating between compression and expansion, then measures what bounces back. These mechanical waves need a physical medium (like body tissue or fluid) to travel through and cannot even propagate in a vacuum.
This fundamental difference is why the FDA actually recommends ultrasound and MRI as alternatives to radiation-based imaging when clinically appropriate. There are no federal radiation safety performance standards for diagnostic ultrasound because it produces no radiation at all.
What Ultrasound Can Do to Tissue
That said, ultrasound is a form of energy and is not completely inert. It can produce two biological effects worth understanding. First, sound waves can cause a slight rise in tissue temperature. Second, they can theoretically create tiny gas pockets in body fluids, a phenomenon called cavitation. Both effects are well understood and carefully controlled during medical scans.
Sonographers monitor two real-time safety indicators on their screens. The thermal index measures how much the sound waves could warm tissue, while the mechanical index tracks the risk of cavitation. Both are designed to stay below 1.0 during obstetric scans. In 1992, the FDA raised the maximum permitted output of ultrasound machines but required that these safety indicators be displayed in real time so operators could adjust their settings.
One concern worth noting: research has found that roughly 20% of operators exceed a thermal index of 1.0 during scans, suggesting that awareness of best practices varies. A study tracking pregnancies where multiple scans exceeded these thresholds found some minor differences in newborn body measurements (slightly smaller limb circumference, small changes in skin fold thickness), but no effects on overall birth size. These findings reinforce why guidelines emphasize keeping exposure as low as possible rather than suggesting any real danger from routine scanning.
No Link to Autism or Developmental Problems
One of the most persistent concerns parents encounter online is whether prenatal ultrasound could increase the risk of autism. A well-designed CDC case-control study of 1,524 singleton pregnancies directly addressed this question. Researchers compared children diagnosed with autism spectrum disorder to controls, looking at how many ultrasounds their mothers received and when. The results were clear: there was no association between the number of prenatal ultrasounds and autism risk, in any trimester, whether or not the mother had medical indications for additional scans. The adjusted odds ratio was essentially 1.0, meaning no increased risk whatsoever.
The study used gold-standard diagnostic assessments for autism classification and carefully accounted for the fact that higher-risk pregnancies tend to receive more ultrasounds in the first place. This is an important distinction, because raw data can make it look like more ultrasounds correlate with problems when the real explanation is that complicated pregnancies need more monitoring and also carry independent risks.
How Many Scans Are Recommended
The World Health Organization recommends at least one ultrasound before 24 weeks of pregnancy. This scan serves several purposes: confirming the pregnancy is in the uterus, checking for a heartbeat, estimating gestational age, determining whether you’re carrying multiples, and screening for structural abnormalities. First-trimester dating by ultrasound is the most accurate method for establishing gestational age, precise to within 5 to 7 days when measuring the embryo’s crown-to-rump length before 14 weeks.
In practice, most people in the U.S. receive at least two ultrasounds during a low-risk pregnancy. The first typically happens in the first trimester to confirm dating and viability, and the second around 18 to 22 weeks for a detailed anatomy scan. Higher-risk pregnancies often require additional monitoring, and the medical benefit of those extra scans is well established. The key principle from every major professional organization (the American Institute of Ultrasound in Medicine, the American College of Obstetricians and Gynecologists, and others) is the same: ultrasound should be performed when there is a valid medical reason, using the lowest possible exposure settings to get the needed information.
The ALARA Principle
Every guideline for obstetric ultrasound references something called ALARA, which stands for “As Low As Reasonably Achievable.” Originally developed for ionizing radiation, this safety framework has been adopted for ultrasound as well. It means that even though diagnostic ultrasound has an excellent safety record, providers should still minimize exposure time and energy output while maintaining image quality.
In practice, this means your sonographer should use the lowest power setting that produces a clear enough image, avoid lingering on one area longer than necessary, and skip unnecessary scanning modes. For example, professional guidelines specifically discourage using pulsed Doppler ultrasound (which delivers more energy than standard imaging) just to listen to or document the embryonic heartbeat in the first trimester when simpler methods work fine. Research has found that most sonographers don’t actively think about ALARA when choosing their output settings, which has prompted the development of automated systems that could adjust power levels based on image quality feedback. In a pilot study, the default machine setting delivered more acoustic energy than necessary for 80% of scans.
Why Keepsake Ultrasounds Are Different
There is one category of prenatal ultrasound that raises legitimate safety concerns: commercial “keepsake” sessions offered by non-medical studios. These businesses use high-powered 3D and 4D equipment to produce videos and photos of the fetus, often for sessions lasting up to an hour. The FDA, the American Institute of Ultrasound in Medicine, and other organizations have all warned against these services.
The concerns are specific and practical. The operators at these studios are typically not trained healthcare providers and may not understand how to monitor thermal and mechanical indices. Modern ultrasound equipment can produce intensities eight times higher than machines from a decade earlier, making operator knowledge more important than ever. Sessions are far longer than diagnostic scans, which increases total energy exposure. And because there is no medical purpose, any risk at all is unjustified.
If you want keepsake images or video of your baby, the FDA’s position is reasonable: those are fine to capture during a medically indicated scan, as long as getting them doesn’t require additional exposure. What you want to avoid is a separate, prolonged session with no diagnostic purpose conducted by someone without medical training.
What This Means for Your Pregnancy
Routine prenatal ultrasound, performed by a qualified provider for a medical reason, has over two decades of safety data behind it and no confirmed evidence of harm to mothers or babies. The sound waves involved carry no radiation, and the biological effects they can produce (mild heating, potential cavitation) are monitored in real time and kept well within safe limits during standard exams. Large studies have found no connection to autism, developmental delays, or meaningful differences in birth outcomes.
The scans your provider recommends are there to catch ectopic pregnancies, confirm your due date, identify twins, screen for structural problems, and monitor growth. The diagnostic value of that information is substantial and well established. Where caution is warranted is in unnecessary, prolonged, or non-medical scanning, particularly from commercial keepsake studios operating outside the medical system.