The Second-Opinion Decision Tree: When a DNA Retest Is Actually Worth It

Running a DNA test a second time is sometimes the right call and sometimes a predictable waste of money. Families who are unsure about a result often assume a second lab will produce a different answer. On a clean 23-marker test, it will not. The second-opinion decision depends entirely on what caused the original result to feel incomplete, and the answer is specific to the situation.

This guide works through the decision tree one branch at a time, so a family holding a confusing or surprising result can figure out which path (if any) actually produces a different outcome.

Branch 1: Did the original test use fewer than 20 markers?

This is the simplest branch and the one with the clearest answer. If the original test used a 13-marker or 16-marker panel (common on legacy pharmacy kits and budget providers), the Combined Paternity Index on a true inclusion is in the thousands rather than millions, and the discriminating power on borderline cases is substantially lower.

If yes: retest on a modern 23-marker panel (or higher). This is one of the few situations where a retest meaningfully changes the information you have. The new result will not necessarily change the answer, but it will produce a stronger version of whatever answer was there.

If no (original was 23+ markers): this branch does not apply. Continue to the next branch.

Branch 2: Could the samples have been handled incorrectly?

Sample-handling errors are the most common source of surprising DNA results, and they are the thing a retest can actually fix (unlike biological results, which it cannot).

Ask:

• Were the sub-envelopes labeled with correct participant names before swabbing?
• Did swabs touch each other during the air-drying step?
• Were swabs sealed before fully drying?
• Did a family member help with collection in a way that could have put their DNA on another person's swab?
• Was each person's mouth free of food, drink, or nursing residue for at least 30 minutes before swabbing?

If any of these are a maybe or yes: retest with a fresh collection at the same lab. Do not switch labs for this; the issue was upstream of the lab, so switching labs does not help. Collecting cleanly this time addresses the actual issue.

If no (collection was clean): continue to the next branch.

The guide on the 48 hours before you ship your DNA samples covers the handling mistakes most commonly responsible for confusing results.

Branch 3: Did the result land in the 0.1 to 10 inconclusive range?

Clean paternity trios essentially never land in the inconclusive middle range on a 23-marker test. If a paternity test produces an inconclusive result, something is wrong (usually in the sample-handling branch above, or occasionally due to closely related participants).

Kinship tests (grandparent, aunt or uncle, half-sibling) regularly land in the inconclusive range. This is normal and reflects the weaker biological signal on kinship questions, not a lab error.

If the inconclusive result was on a kinship test:

• Confirm the lab's reflex-marker protocol was run. USDC runs additional autosomal markers (up to 28 total) on the same sample when the first-pass 23-marker result is inconclusive. If the reflex was not run, that is the first step.
• If reflex markers were already run and the result is still inconclusive, adding a second biological relative to the testing is more productive than retesting the same two people. The guide on kinship DNA testing when only one relative is available covers which relative to add.

If the inconclusive result was on a paternity trio:

• Check sample handling (Branch 2). This is almost always where the issue is.
• Check whether the mother and alleged father are closely related. If yes, flag it to the lab.
• Retest with a fresh collection. The result should not land in the inconclusive range on a second clean try.

Branch 4: Does the result need to hold up in court?

A home test produces a scientifically accurate result but no documented chain of custody. If the result needs to be used in a court proceeding, government filing, or similar formal use, the home test is not sufficient, regardless of how clean the lab result is.

If yes (legal use): order a chain-of-custody collection, which uses the same lab science but adds the identity verification, witnessed collection, and documented handoff that courts require. The result will almost certainly match the home test, but it will come with the documentation that makes it admissible. The guide on what chain of custody actually adds to a legal DNA test covers this distinction.

If no (personal knowledge only): this branch does not apply.

Branch 5: Does the result just feel wrong, with no specific concern?

A result that is surprising or hard to accept, but came from a clean 23-marker test with correct sample handling, is not likely to change with a retest. The multiplicative math behind the Combined Paternity Index produces decisive results, and running the same test again with the same methodology produces the same number.

Families in this situation often retest hoping for a different answer. They almost always get the same answer, which costs them time and money without resolving the discomfort.

What actually helps instead:

• Read the report carefully, including the per-marker breakdown. Understanding where the number comes from often resolves the "it doesn't feel right" feeling better than another test.
• Call the lab and ask them to walk through the result. Reputable labs have customer service that will explain the math on a specific report.
• Consider whether the discomfort is really about the test or about what the result implies for family relationships. A therapist can help with the latter; another test cannot.
• Give it time. A result that feels wrong on day one often settles into "accurate but still hard" over weeks. That is a different problem than "possibly inaccurate" and does not need another test.

Decision summary

Running through the branches:

1. Original test used fewer than 20 markers: yes, retest on a 23-marker panel.
2. Sample handling could have been compromised: yes, retest with clean collection at the same lab.
3. Result was inconclusive on a kinship test: yes, add a second biological relative (or confirm reflex markers were run).
4. Result needs to be court-admissible: yes, order chain-of-custody collection.
5. Result just feels wrong but no specific concern: no, retesting will produce the same answer. Address the discomfort differently.

Families who work through this tree before ordering a second test save money, time, and the emotional overhead of running another round of a test that was going to produce the same answer. For situations where retest is the right call, choosing the specific thing to change (panel size, sample handling, biological relative added, chain of custody) matters more than just running the test again.

What if two retests disagree?

Occasionally families end up with two results that seem to disagree. Before concluding that one lab is wrong, check:

• Did both tests use the same marker count? A 13-marker test and a 23-marker test can produce different-feeling results on borderline cases without either being wrong.
• Did both tests use clean samples with proper handling?
• Are the actual values on the two reports really contradictory, or are they both in the same categorical range (both supportive, both excluding, both inconclusive)?

The guide on why two home DNA tests can give different results covers the three real reasons two tests can produce different outcomes, which are almost always resolvable without a third test.

USDC's home paternity test ($79) uses the 23-marker standard panel with reflex up to 28 markers on inconclusive kinship cases. For most families, the first test is also the last test.