For every 1,000 people invited to long-term PSA screening, the best current evidence suggests about two fewer deaths from prostate cancer—and about 36 more prostate-cancer diagnoses.

Those two numbers explain why a prostate-specific antigen, or PSA, blood test has never fitted neatly into “screen” or “do not screen”. Finding cancer earlier can prevent some metastatic disease and deaths. It can also find slow-growing cancers that would never have caused illness, exposing people to years of surveillance or treatment they did not need.

Australia’s new 2026 guideline does not remove that trade-off. It changes the pathway around the test: informed discussion before screening, repeat testing after an elevated result, risk-adjusted thresholds, prostate MRI before biopsy and active surveillance for many low-risk cancers.

What PSA can—and cannot—show

PSA is a protein made by prostate cells. Cancer can raise its concentration in blood, but so can benign prostate enlargement, inflammation, infection, recent procedures and other non-cancer causes. Some clinically important cancers produce a PSA below the usual action level.

A PSA result is therefore a risk signal, not a diagnosis. Screening refers to testing someone without symptoms. A person with symptoms or a clinical finding that raises concern is being assessed through a different pathway, even if PSA is one of the tests used.

Screening reaches people who feel well. Any benefit may take well over a decade to appear, while an abnormal result can start a diagnostic sequence immediately.

The mortality benefit is real but small

A 2026 Cochrane review combined six randomised trials involving 789,086 participants aged 45 to 80. Screening schedules and PSA thresholds differed, and follow-up ranged from 3.2 to 23 years.

The review’s most reliable mortality estimate came from the large European Randomized Study of Screening for Prostate Cancer. Over 23 years, prostate-cancer deaths were estimated to fall from 16 per 1,000 without organised screening to 14 per 1,000 with screening. The rate ratio was 0.87, with a 95% confidence interval from 0.80 to 0.95.

The same evidence estimated 36 additional prostate-cancer diagnoses per 1,000 people screened, including 34 additional localised cancers. Metastatic diagnoses fell by about five per 1,000. The review did not establish a reduction in overall mortality.

A separate UK trial, CAP, randomised 408,721 men aged 50 to 69 through their general practices. A single invitation to PSA testing reduced prostate-cancer mortality at 15 years from 0.78% to 0.69%—an absolute difference of 0.09 percentage points. Overall mortality was virtually identical.

CAP tested one invitation, and only about 40% of invited men attended the PSA clinic. Some men in the control group obtained PSA tests anyway. It is not a direct test of Australia’s current two-yearly approach, but it shows how a statistically reliable relative reduction can translate into a small absolute difference.

The Göteborg trial tested repeated screening more intensively. It randomised almost 20,000 Swedish men aged 50 to 64 to biennial invitations or no invitation. After 22 years, prostate-cancer mortality was 1.55% in the screening group and 2.13% in controls. Prostate cancer was diagnosed in 15.4% and 11.1% respectively. Screening shifted both outcomes: fewer prostate-cancer deaths and substantially more diagnoses.

More diagnoses are not automatically more health

Overdiagnosis means finding a cancer that would not have caused symptoms or death during the person’s lifetime. It cannot usually be identified with certainty at the moment of diagnosis. The concern is not that the pathology is imaginary; it is that its discovery may not improve the person’s future.

Older screening trials often moved from an elevated PSA to systematic biopsy, then from a cancer diagnosis to surgery or radiotherapy. Biopsy can cause bleeding, pain, infection and, occasionally, hospital admission. Treatment can affect urinary continence, erections and bowel function. The 2026 Cochrane review found that trial reporting of these harms was too incomplete to produce a dependable modern estimate.

Clinical practice has changed. Multiparametric MRI combines several magnetic-resonance sequences to assess whether a suspicious area is present before biopsy. In a Swedish randomised screening study, 1,532 participants with PSA at or above 3 micrograms per litre entered either a standard-biopsy pathway or an MRI pathway. Clinically significant cancer was found in 21% of the MRI group and 18% of the standard group. Clinically insignificant cancer was found in 4% and 12% respectively.

That study measured detection in the first screening round, not deaths over decades. MRI can reduce avoidable biopsy and low-grade diagnoses; it cannot guarantee that an important cancer will never be missed. Active surveillance has also expanded, allowing many low-risk cancers to be monitored rather than treated immediately. Surveillance reduces overtreatment, though the diagnosis and repeated follow-up still carry physical and psychological costs.

What Australia’s 2026 guideline recommends

The new national guideline was approved by the National Health and Medical Research Council in May 2026. It recommends that testing follow a discussion of possible benefits and harms.

For people at average risk who decide to test, the guideline recommends PSA every two years from age 50 to 69. A result at or above 3.0 micrograms per litre should be repeated within one to three months before referral is offered. Higher-risk groups—including some people with a strong family history, Black sub-Saharan African ancestry or a BRCA2 mutation—are offered earlier testing from age 45 and lower action levels.

From age 70, the decision is based on clinical assessment, other health conditions, preferences and life expectancy, generally more than seven years. A persistently elevated result leads to specialist assessment, with multiparametric MRI recommended before any biopsy.

These are recommendations for planned, informed testing, not an Australian population screening program. A threshold is an action point for repeat assessment, not a line separating cancer from no cancer.

The question before the blood test

PSA screening works over a long horizon. Its possible benefit depends on age, inherited and family risk, general health, willingness to proceed through MRI or biopsy, and how a low-risk diagnosis would be managed. Starting with the blood test and discussing consequences only after it is elevated reverses the order of informed decision-making.

A useful clinical conversation asks: if this result is high, what would happen next—and would I want to follow that pathway? The answer can include repeat PSA, risk assessment, MRI, targeted biopsy, active surveillance or treatment. Each step changes the chance of benefit and harm.

The new Australian pathway is more selective than the one tested in the original trials. That should reduce avoidable procedures and treatment, but it has not yet been followed long enough to show its own effect on mortality. The honest case for PSA screening is neither a simple life-saving test nor a useless source of overdiagnosis. It is a small long-term chance of avoiding metastatic disease or prostate-cancer death, exchanged for a much larger chance of entering a diagnostic pathway.

Sources

Evidence note: This Insight is general education, not individual medical advice. It does not recommend whether a particular person should have PSA testing or what should be done with a result. Screening choices and abnormal results should be discussed with a qualified clinician.