The radiation from a CT scan is "ionising" and can raise the risk of cancer slightly; but the increase in risk from one CT scan is small, and when the test is needed to diagnose cancer or another serious illness, its benefit far outweighs this risk [1]. For comparison, a standard chest CT scan gives about as much radiation as 2 years of natural background radiation; plain X-rays and mammograms give much less, and MRI and ultrasound use no ionising radiation [1, 11, 12]. Children are more sensitive to radiation, so it is important that a test is done only when needed and at the lowest suitable dose [1, 7].
Key points
- All of us receive radiation every year from the ground, the air (radon gas) and space; on average about 3 millisieverts (mSv) a year [1].
- A plain chest X-ray gives very little radiation, equal to a few days of natural background radiation; a CT scan usually equals a few months to a few years [1, 2].
- The risk from each CT scan is small for each person: for a CT scan with a fairly high dose, about 1 person in 2,000; this estimate is not certain [2].
- MRI and ultrasound use no ionising radiation and can sometimes replace a test that uses radiation [11, 12].
How much radiation does each test give?
Radiation is measured in a unit called the "millisievert" (mSv). The simplest way to understand these numbers is to compare them with the natural radiation all of us receive from our surroundings each year: in the United States, about 3 mSv a year on average [1]. The actual dose of each test varies with the machine, body size and type of test [2, 3].
| Test | Radiation (mSv) | About the same as natural background radiation for |
|---|---|---|
| Natural background radiation | 3 per year [1] | 1 year |
| Plain chest X-ray (from the front) | 0.02 [2] | A few days |
| Plain X-ray of the lower spine | 1.5 [2] | 6 months |
| Low-dose lung CT (for screening) | 1.5 [1] | 6 months [1] |
| Head CT | 2 [2] | 8 months |
| Standard chest CT | 6.1 [1] | 2 years [1] |
| Abdomen and pelvis CT | About 6 [1] | 2 years |
| Heart CT angiography | 16 [2] | More than 5 years |
| Whole-body CT | 10 to 20 [1] | More than 3 years [1] |
| MRI and ultrasound | No ionising radiation [11, 12] | — |
These are average values; the actual dose of a test can be 2 to 3 times lower or higher and varies widely between centres [2, 3].
How risky is a CT scan?
The radiation from a CT scan is "ionising", a type of radiation that can raise the risk of cancer [1]. A CT scan gives more radiation than a plain X-ray, but the increase in cancer risk from one CT scan is small [1].
The US Food and Drug Administration estimates that a CT scan with a fairly high dose (10 mSv) may lead to a fatal cancer in about 1 person in 2,000; by comparison, in the United States about 1 person in 5 develops a fatal cancer during their lifetime from other causes [2]. These estimates are not certain and rest on the cautious assumption that any amount of radiation, however small, carries some risk [2].
The risk is slightly higher for women than for men [1], and the younger the person, the higher it is: in one analysis, the risk for 20-year-olds was about twice, and for 60-year-olds about half, that for 40-year-olds [3]. This small risk becomes important for the population when very large numbers of CT scans are done without good reason [2, 8].
Children and teenagers
Children are more sensitive to radiation, because their bodies are growing, their cells divide faster and they have more years of life ahead of them [1]. Several large studies have followed children and young people who had CT scans for many years [4, 5, 6, 7]:
- In Britain: the higher the accumulated dose, the more leukaemia and brain tumours there were; but because these cancers are rare, in the 10 years after the first scan, about one extra case of leukaemia and one extra brain tumour were estimated for every 10,000 head CT scans in children under 10 [4].
- In 9 European countries, nearly one million people: for every 10,000 children scanned today, 1 to 2 are expected to develop a blood cancer caused by the radiation in the following 12 years [7]; for brain cancer, too, the risk rose with the dose [6].
So the risk is real but small. If the test is truly needed, its benefit for the child is greater than this small risk; what matters is that CT scans are not done without reason or repeated unnecessarily, that the dose is adjusted to the child's size, and that a radiation-free test is chosen if it is just as useful [1, 4, 7].
Mammograms and plain X-rays
Plain X-rays, such as a chest X-ray, and mammograms give relatively little radiation [1]. It is estimated that one mammogram at age 40 leads to a fatal breast cancer in about 1 in every 100,000 women [10]. By contrast, yearly mammograms from age 40 to 74 prevent about 60 deaths from breast cancer through early detection for every death the radiation might cause [9]. So worry about radiation should not stop you from having mammograms on time; see At what age should I start mammograms?
MRI, ultrasound and PET scans
MRI makes images with a strong magnet and radio waves and uses no ionising radiation [11]. Ultrasound works with sound waves and also uses no ionising radiation [12]. The US Food and Drug Administration recommends considering these tests instead of tests that use radiation when they are suitable [12]; but they do not always give the same information: for example, MRI shows soft tissue better, while CT usually shows bone better [11]. Choosing the right test is a decision for your doctor.
A PET scan, which is usually done together with a CT scan, gives more radiation [1]; why and when it is used is explained in What is a PET scan and how is it used in cancer?
What should I do? Ask these three questions before a CT scan
If your doctor has recommended a CT scan or an X-ray, do not put it off because of worry about cancer; when the test is needed to look into symptoms or a serious illness, its benefit far outweighs its risk [1]. You can ask these three questions [1]:
- Why is this test needed?
- Will the result change my treatment?
- Is there another test without radiation that would be just as useful?
If you recently had a CT scan somewhere else, bring the images and the report so that the test is not repeated without reason. If you could be pregnant, be sure to tell the doctor and the radiographer [1].
Frequently asked questions
How many CT scans a year are safe?
I have cancer and have CT scans regularly; should I worry?
Is a CT scan risky in pregnancy?
If you could be pregnant, be sure to say so; the doctor may lower the radiation dose or choose another test. According to the US National Cancer Institute, the radiation from a CT scan is believed to be too low to harm a developing baby [1]. Ultrasound uses no ionising radiation and is useful for women of childbearing age [12].
Is a plain chest or bone X-ray risky?
Is a whole-body CT scan a good idea for a check-up?
A whole-body CT scan gives 10 to 20 mSv, more than three times the natural background radiation of a year [1]. The US Food and Drug Administration warns that a scan without a clear reason may find harmless things that lead to further, sometimes invasive, tests [2]. For screening, have the tests recommended for your age and situation: cancer screening.
Aren't yearly mammograms risky?
In one analysis, yearly mammograms from age 40 to 74 prevented about 60 deaths from breast cancer for every death the radiation might cause; the radiation risk was 5 times lower in women screened every 2 years from age 50 [9]. Decide when to start and how often with your doctor: At what age should I start mammograms?
Notes for doctors (for physicians)
For physicians
Doses. FDA (average effective dose, based on McCollough 2015): PA chest radiograph 0.02, lumbar spine 1.5, head CT 2, chest CT 7, abdominal CT 8, coronary calcium CT 3 and coronary CT angiography 16 mSv; the actual dose may be 2 to 3 times higher or lower. FDA estimate for a 10 mSv CT: an increase in the chance of fatal cancer of about 1 in 2,000, based on the linear no-threshold model [2]. NCI: low-dose lung CT 1.5, standard chest CT 6.1, CT colonography and abdomen–pelvis CT about 6, whole-body CT 10 to 20 and PSMA PET/CT 4 to 23 mSv; natural background about 3 mSv a year [1]. Smith-Bindman 2009 (1,119 adult patients at 4 centres): median effective dose from 2 mSv (routine head CT) to 31 mSv (multiphase abdomen–pelvis CT), with a mean 13-fold variation between the highest and lowest dose for each study type; lifetime attributable risk of cancer from coronary CT angiography at age 40, 1 in 270 women (1 in 600 men), and from routine head CT 1 in 8,100 women (1 in 11,080 men); about double at age 20 and about 50% lower at age 60 [3].
Cohorts of children and young people. Pearce (Lancet 2012; Great Britain, CT before age 22 in 1985 to 2002; 74 leukaemias among 178,604 and 135 brain tumours among 176,587 patients): ERR per mGy for leukaemia 0.036 (0.005 to 0.120) and brain tumours 0.023 (0.010 to 0.049); RR of leukaemia with cumulative red bone marrow dose of 30 mGy or more 3.18 (1.46 to 6.94) and of brain tumours with 50 to 74 mGy 2.82 (1.33 to 6.03) vs less than 5 mGy; one excess leukaemia and one excess brain tumour per 10,000 head CTs before age 10 over 10 years [4]. Mathews (BMJ 2013; 680,211 exposed people in an Australian cohort of 10.9 million): IRR for all cancers 1.24 (1.20 to 1.29), rising by 0.16 per additional CT; 608 excess cancers; absolute excess incidence 9.38 per 100,000 person-years; mean effective dose per scan 4.5 mSv [5]. EPI-CT (9 European countries): brain cancer in 658,752 people (165 cases), ERR per 100 mGy brain dose 1.27 (0.51 to 2.69) [6]; haematological malignancies in 948,174 people (790 cases), ERR per 100 mGy bone marrow dose 1.96 (1.10 to 3.12); at today's mean dose of 8 mGy, 1 to 2 radiation-attributable cases per 10,000 children over 12 years [7]. To reduce reverse causation, follow-up started after a 2- to 5-year lag, and sensitivity analyses (longer lag, excluding possibly previously unreported cancers) did not change the results [4, 6].
Population projection. Smith-Bindman (JAMA Intern Med 2025; BEIR VII models): 93 million CTs in 61.5 million US patients in 2023 may cause about 103,000 lifetime cancers (91% in adults; the largest share from abdomen–pelvis CT); if current practice continues, up to 5% of annual cancer diagnoses. This is a model-based projection, not a direct observation [8].
Mammography. Hendrick (Radiology 2010): mean glandular dose of two-view digital mammography 3.7 mGy; lifetime attributable risk of fatal breast cancer 1.3 per 100,000 women aged 40 [10]. Miglioretti (Ann Intern Med 2016): annual screening from 40 to 74 in 100,000 women, 125 radiation-induced breast cancers (88 to 178) and 16 deaths (11 to 23) vs 968 deaths averted; in women with large breasts needing extra views (8%), 266 cancers and 35 deaths; biennial screening from age 50 reduced the risk 5-fold [9].
In practice. Justify each test (will the result change management?), use protocols adapted to patient size, especially in children, and consider ultrasound or MRI when suitable; the ALARA principle [1, 12].
References
12 referencesShow all referencesClose the reference list
Numbers in the text, such as [1], refer to these sources. Sources with a “Study card” button are summarised in the site’s evidence library.
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
This page is based on reputable scientific sources and was last updated on 28 October 2026. The content of this website is for educational and informational purposes only and does not replace consultation, diagnosis or treatment by a specialist physician. Consult your treating physician about any treatment decision.
Printed from the website; the full and up-to-date version: https://drdabiri.com/en/questions/do-ct-scans-cause-cancer/