How to track your PSA over time

PSA is a test where the direction matters as much as the number. A result of 3.1 that was 1.4 three years ago means something different from a 3.1 that has sat between 2.8 and 3.4 for a decade. If you are on active surveillance, or have been treated for prostate cancer, the whole point of testing is to watch how PSA changes. A careful history, drawn and recorded the same way each time, tells you more than any single report.

What PSA measures

PSA stands for prostate-specific antigen. The National Cancer Institute describes it as a protein made by both normal and cancerous cells of the prostate. A small amount reaches the blood. Cancer can raise it, but so can a larger prostate or an infection. The test is about the prostate, not specifically about cancer. A report can show a few related values:

• Total PSA: the standard test, usually just labeled “PSA”. All the thresholds below refer to it.
• Free PSA: the part of PSA that is not bound to proteins in the blood.
• Percent free PSA: free PSA divided by total PSA, times 100. A lower percentage points more toward cancer. In a 1998 multicenter study of 773 men with a total PSA of 4 to 10, the chance of cancer on biopsy was about 56 percent when free PSA was 10 percent or less, and under 10 percent when it was above 25 percent.
• PSA density: total PSA divided by prostate volume from an MRI or ultrasound. It appears in imaging reports more than lab reports.

The numbers on the report

PSA is reported in ng/mL in the United States and in µg/L in much of the rest of the world. They are the same number, so 2.4 ng/mL is 2.4 µg/L.

There is no single normal

A PSA above 4.0 has long been treated as abnormal, but the NCI fact sheet says plainly that no single threshold separates a normal PSA from an abnormal one. The AUA/SUO early detection guideline (2023, amended 2026) explains that 4.0 came from very early studies of men thought to be cancer-free, and that 3.0 was the biopsy threshold in the Finnish arm of a large European screening trial. Because PSA rises with age even without cancer, the guideline says 4.0 is too high for men in their 40s and 50s and too low for men in their 70s and 80s. The age-specific values it cites:

The USPSTF, whose 2018 recommendation is still current while an update is in progress, notes that screening trials used biopsy thresholds from 2.5 to 10.0 ng/mL, and that lower ones bring more false alarms and biopsies.

Both treat a high PSA as the start of a decision, not a diagnosis. The AUA guideline suggests risk calculators, MRI, or other markers before a biopsy, and your clinician may use a different threshold for you based on family history or a prior biopsy.

How often it is tested

• Screening: the AUA guideline suggests a baseline PSA between ages 45 and 50, or 40 to 45 for men at higher risk (Black men, men with certain inherited mutations, and men with a strong family history), then every 2 to 4 years from 50 to 69. The USPSTF leaves screening at 55 to 69 to an individual decision and recommends against it at 70 and older.
• Active surveillance: the AUA/ASTRO guideline on localized prostate cancer (2022, amended 2026) monitors with serial PSA tests and repeat biopsies, and advises PSA generally no more often than every six months.
• After surgery or radiation: the same guideline suggests PSA every 3 to 6 months for the first two years, every six months to year five, and yearly after that, with the schedule tailored to your risk.

What counts as a real change

Repeat a high result before acting

The first rule of reading PSA is that one high value is not enough. The AUA early detection guideline recommends repeating a newly elevated PSA before any further markers, imaging, or biopsy. It notes that 25 to 40 percent of newly elevated PSA results return to normal on retesting, and it suggests a repeat in a few months. On active surveillance, the localized cancer guideline likewise says a rise should first prompt a repeat, because short-lived elevations are common.

How much PSA wobbles on its own

In the European Biological Variation Study, 37 healthy men aged 22 to 59 were tested weekly for ten weeks. Total PSA varied by about 7 percent within each man, and the authors put the change needed to be confident of a real rise at 15 to 20 percent. That is healthy men under stable conditions, so real-life noise is larger. A move from 2.0 to 2.3 is not news by itself, but if your last five results ran from 1.8 to 2.2, a 3.4 deserves a repeat even though it is under 4.0. Several results moving the same way say more than one, which is the idea behind velocity and doubling time.

PSA velocity and doubling time

Velocity

PSA velocity is the rise per year, in ng/mL per year. A PSA that went from 2.0 to 2.6 over 18 months has a velocity of 0.4 ng/mL per year.

Velocity was once suggested as a reason for biopsy on its own. It did not hold up. In a study of 5,519 men from the Prostate Cancer Prevention Trial who all had biopsies, adding velocity barely improved the prediction of cancer, and biopsying men for high velocity alone would mean biopsying about one man in seven. The authors concluded that velocity should not be in practice guidelines. The AUA guideline now recommends against using PSA velocity as the only reason for further testing or biopsy during screening. It also notes that a very fast rise, over 3 ng/mL per year, is more often linked to inflammation than to cancer. In screening, a rising line is a reason to talk to your clinician, not a verdict.

Doubling time

PSA doubling time is how long PSA takes to double. It assumes PSA grows by a steady percentage. It matters most after treatment. In a study of 379 men whose PSA came back after surgery, a shorter doubling time was one of three factors that predicted death from prostate cancer, alongside Gleason score and how soon the PSA came back. It helps your care team judge how quickly a returning PSA needs attention. On active surveillance, PSA changes count alongside MRI and biopsy, not on their own.

The PSA Working Group set out how to calculate it:

1. Take the natural log (ln) of each PSA value.
2. Fit a straight line through ln(PSA) against time (a linear regression).
3. Doubling time = ln(2) ÷ the slope of that line.

The same group set minimums for a meaningful result: at least three PSA values, at least four weeks apart, spanning at least three months, all from the same assay and preferably the same lab. Each value should be 0.20 ng/mL or higher, and all values from up to 12 months should be included, not just the ones that rose.

A worked example. Suppose PSA after surgery reads 0.20 at month 0, 0.28 at month 4, and 0.40 at month 8.

The slope of ln(PSA) against months is about 0.0866 per month. Doubling time is 0.693 ÷ 0.0866, or about 8.0 months. That matches the eye: PSA went from 0.20 to 0.40 in eight months.

The PSA doubling time calculator does this from your dates and values, gives velocity too, and checks the Working Group minimums. In a spreadsheet, with dates in column A, PSA in column B, and =LN(B2) in column C, =LN(2)/SLOPE(C2:C4, A2:A4) gives the doubling time in days. The lab results spreadsheet template is a starting point. Use either to follow along, not to replace your care team’s numbers.

Things that move PSA

Several things change PSA without saying anything about cancer:

• Infection and inflammation. The AUA guideline lists urinary tract infections as a cause of short-lived rises, and the NCI adds prostatitis. The guideline also supports the Choosing Wisely advice not to give antibiotics for a high PSA in a man without symptoms.
• Instrumentation. A catheter, cystoscopy, biopsy, or urinary retention can raise PSA. PSA has a half-life of 2 to 3 days, and the guideline advises waiting long enough for it to return to baseline before repeating the test.
• Ejaculation, cycling, and a rectal exam. The AUA guideline says that a digital rectal exam and bike riding do not appreciably change PSA, and that most controlled studies of ejaculation found no effect or a modest rise of about 10 percent. The NCI takes a cautious line and suggests avoiding ejaculation and vigorous cycling for two days before a test, which costs little and removes a variable.
• A larger prostate. The NCI names benign enlargement (BPH) in particular among the non-cancer causes of a higher PSA. PSA density adjusts for gland size.
• Finasteride and dutasteride. These drugs, used for BPH and hair loss, lower PSA. The Proscar label says finasteride cuts PSA by about half within six months, and the Avodart label says the same for dutasteride within 3 to 6 months. Both say to double an isolated PSA for comparison with the usual ranges, to set a new baseline once treatment has settled, and that any confirmed rise from the lowest value on treatment may signal cancer. Percent free PSA is unchanged. The hair-loss dose counts too: in a trial of finasteride 1 mg, median PSA fell 40 to 50 percent after 48 weeks. The AUA guideline adds that the drop varies from man to man, so doubling is a rough guide.
• Testosterone therapy. The Endocrine Society guideline checks PSA before treatment and again 3 to 12 months after starting. In that first year, it recommends a urology referral for a confirmed rise of more than 1.4 ng/mL above baseline, or a confirmed PSA above 4.0. See How to track your testosterone over time.
• A change of lab. PSA tests can be calibrated to different standards. In a 2008 study, the same blood measured about 20 percent lower when the test was calibrated to the WHO standard than to the older Hybritech standard. A PSA that drops from 3.0 to 2.4 when you switch labs may be the calibration. Why are lab reference ranges different? explains more about why labs disagree.

After treatment

After surgery to remove the prostate, PSA should fall to an undetectable level, since the gland that made it is gone. An AUA panel defined biochemical recurrence after surgery as a PSA of 0.2 ng/mL or higher, confirmed by a second test above 0.2. If your results after surgery come from a test with a different name, such as an ultrasensitive PSA, keep them on the same history and note the switch.

After radiation, the prostate is still in place, and recurrence is judged differently. The Phoenix definition treats a rise of 2 ng/mL or more above the lowest value after treatment (the nadir) as biochemical failure. That makes your lowest value the number to keep track of.

These definitions are for your care team to apply, but a history with the nadir and treatment dates makes that conversation easier.

What to record with each result

For each PSA, keep:

PSA goes by several names on reports: PSA, Prostate Specific Antigen, PSA, Total, PSA, Ultrasensitive, and PSA-Prostate Specific Antigen, Total, among others. Free PSA and percent free PSA are separate tests. Merge the total PSA names into one series and keep free PSA on its own line.

Charting it

A few choices make a PSA chart honest:

• Space points by date. Velocity and doubling time depend on real time gaps.
• Mark medication starts and stops. A drop of about half in the six months after starting finasteride is expected, and the new baseline starts there.
• Mark lab changes. A 20 percent step at a switch of labs may be calibration.
• After treatment, mark the date of surgery or radiation and the nadir.
• Consider a log scale for doubling. On a log scale, PSA that doubles at a steady rate is a straight line, and a steeper line means a shorter doubling time.

How to track your lab results over time has the general method for any test. Did the medication actually do anything? covers reading a before-and-after fairly.

Tracking PSA in HealthViewer

HealthViewer is a lab results tracker that runs on your own computer. It imports PSA results from lab report PDFs, MyChart downloads, and direct connections to your health system. It files PSA, Prostate Specific Antigen, and PSA-Prostate Specific Antigen, Total under one test, along with total PSA sent by your health system under its standard code. Free PSA from a direct connection files as its own test, Free PSA. Results in ng/mL and µg/L share one chart. Other names, such as “PSA, Total” or an ultrasensitive PSA, arrive as separate tests. Combine the total PSA names with Merge… in the Test Catalog, and keep free PSA separate.

There is no built-in tag for PSA, so add your own in the Test Catalog to PSA, free PSA, and testosterone if you take it, then filter by it in Trends. Choose the Timeline x-axis so points are spaced by date. The overlay view puts total and free PSA on one chart. If your lab reports free PSA but not the percentage, you can add a derived test in Settings with the formula A / B * 100, where A is Free PSA and B is PSA, and a date window of 0 so only same-day results are paired. Trends charts PSA on a plain scale, and HealthViewer does not calculate velocity or doubling time, so work those out with the calculator or as shown above.

Medication markers show where finasteride, dutasteride, or testosterone started and stopped on the PSA chart. Enter each dose as its own period in Medications to mark dose changes too. The journal holds the rest: the date of a biopsy, a catheter, a UTI, surgery or radiation, or a change of lab. Open the Journal menu in Trends and choose Show journal to see those notes as markers on the chart. If your clinician uses an age-adjusted threshold for you, set it as a custom range in the Test Catalog, and the chart shades your range instead of the lab’s.

Everything stays in an encrypted file on your computer, with no account and no cloud service. Getting Started walks through the first import.

This post is general information, not medical advice. Talk to your clinician about what your PSA results mean for you and before starting, stopping, or changing any treatment.