The CEA blood test is not accurate enough to screen for or diagnose cancer on its own, but it performs reasonably well as a monitoring tool for people already diagnosed with certain cancers, particularly colorectal cancer. Its usefulness depends entirely on what it’s being used for. As a one-time screening test in a healthy person, it misses too many cancers and flags too many non-cancerous conditions to be reliable. As a recurring test that tracks changes over time in a cancer patient, it becomes significantly more informative.
Why CEA Fails as a Screening Test
CEA, short for carcinoembryonic antigen, is a protein that some cancers produce in elevated amounts. The problem is that many cancers that typically raise CEA levels don’t always do so. You can have colorectal, lung, pancreatic, or breast cancer and still get a completely normal CEA result. This means the test misses real cancers, which in medical terms is called poor sensitivity.
At the same time, plenty of non-cancerous conditions push CEA levels up. Smoking is the most common culprit, but inflammatory bowel disease, liver disease, chronic lung conditions, and infections can all elevate CEA. This creates false alarms, where the test looks abnormal but no cancer exists. Because of both problems, no major medical organization recommends CEA as a cancer screening tool for the general population.
How Accurate It Is for Monitoring Recurrence
Where CEA testing shows real value is in tracking colorectal cancer patients after treatment. A large review by the UK’s National Institute for Health and Care Research pooled data from dozens of studies and found that accuracy shifts depending on where you set the cutoff for an abnormal result.
At a cutoff of 2.5 ng/mL, the test correctly identified about 82% of recurrences and correctly ruled out recurrence about 80% of the time. That’s a reasonable detection rate, but one in five cancer-free patients would get a false alarm. At a cutoff of 5 ng/mL, detection dropped to 71%, but the false alarm rate improved, with 88% of cancer-free patients correctly identified. Raising the bar to 10 ng/mL caught only 68% of recurrences but was correct 97% of the time when it said a patient was clear.
In practical terms, this means lower cutoffs catch more recurrences but trigger more unnecessary worry and follow-up testing. Higher cutoffs miss more recurrences but give you greater confidence when the result is normal. Your oncologist chooses the threshold based on your individual risk profile and how aggressively they want to surveil.
Trends Matter More Than Single Results
A single CEA number, taken in isolation, tells you relatively little. What matters is the pattern over time. Mayo Clinic Laboratories notes that single values of CEA are less informative than changes assessed over time. A CEA level that steadily climbs across several tests is a much stronger signal of recurrence than one elevated reading, which could reflect a temporary infection, a lab variation, or even recent chemotherapy.
This is why oncologists establish a baseline CEA level before treatment begins. After surgery or chemotherapy, they expect CEA to drop. If it does and then stays low across multiple checks, that’s reassuring. If it rises again over consecutive tests, that pattern prompts imaging or further investigation. The American Society of Clinical Oncology recommends CEA testing every three months for at least three years after treatment in patients with stage II or III colorectal cancer who would be candidates for additional treatment if recurrence were found.
One important timing detail: certain chemotherapy drugs can temporarily raise CEA levels on their own. Doctors typically wait until adjuvant treatment is finished before starting routine surveillance to avoid misleading readings.
What Normal CEA Levels Look Like
For most people, a normal CEA level falls between 0 and 3 ng/mL. If you smoke, levels up to 5 ng/mL are generally considered within the normal range. UCSF Health uses a slightly tighter reference of 0 to 2.5 ng/mL for non-smokers, so the exact cutoff can vary slightly between labs.
An elevated CEA in someone with no cancer history doesn’t automatically point to cancer. It’s a nonspecific marker, meaning it responds to inflammation and tissue damage from many sources. Your doctor would consider the result alongside your symptoms, medical history, and other test results before drawing any conclusions.
What Happens After an Elevated Result
If your CEA comes back high, the next step is almost never a cancer diagnosis based on that number alone. For cancer patients in surveillance, a rising CEA typically triggers imaging, such as a CT scan, to look for visible evidence of recurrence. The CEA result is a flag that says “look closer,” not a definitive answer.
For people without a cancer history who happen to have an elevated CEA, one study found that up to 7.4% eventually developed a malignancy within two years. That means the vast majority of elevated results in this group turn out to be caused by something other than cancer. Still, an unexplained high result usually warrants follow-up testing and monitoring over the next one to two years to make sure nothing is developing.
The bottom line is that CEA works best as one piece of a larger surveillance puzzle. It’s a useful early warning system when tracked over time in cancer patients, but it’s not precise enough to stand on its own as either a screening tool or a diagnostic test.