The Short Answer
A PIP descriptor applies where it is satisfied on more than half of the days in the period being assessed. GOV.UK states the test as expecting "to have difficulty more than half of the days over a 12-month period". The rule counts days rather than averaging severity, so a descriptor applying on 4 days in every 7 applies for the whole period. Good days do not dilute it into a lower descriptor.
What the 50% Rule Says
More than half the days, counted per activity. The rule exists because most conditions vary, and without it a fluctuating condition would score whatever it happened to be doing on the day of the assessment.
GOV.UK sets out the test in the eligibility section: "To get PIP you must experience difficulty with tasks most of the time. This means that you expect to have difficulty more than half of the days over a 12-month period."
Last checked: 7 August 2026
The regulations apply it to each of all 12 PIP descriptors separately. A claim can satisfy the rule on washing and dressing while failing it on cooking, and the score reflects that activity by activity.
How to Count the Days
Take a normal week or month and count, rather than estimating a percentage. The arithmetic is simple once the period is real rather than notional.
| Pattern | Days the descriptor applies | Does it apply? |
|---|---|---|
| 4 days in every 7 | 57% | Yes |
| 3 days in every 7 | 43% | No |
| 2 weeks bad, 2 weeks manageable | 50% | Not on its own — see below |
| Every day but two a month | 93% | Yes |
Last checked: 7 August 2026
Exactly half is not more than half, so a strict 50/50 split does not satisfy the rule for a single descriptor. What resolves that case is the combination rule below.
Where Two Descriptors Share the Days
Descriptors in the same activity can be added together to clear the threshold. Regulation 7 of the PIP Regulations 2013 provides for this, and where the descriptors apply equally often the higher-scoring one is scored. Where no single descriptor applies on more than half the days, but two or more from one activity together do, the rule looks at them jointly.
The one that applies on the most days is the one scored. Where they apply on the same number of days, the higher-scoring one is scored. So a person who needs an aid to wash on 3 days a week and assistance to wash on 3 more satisfies the activity on 6 days out of 7, and scores the assistance descriptor as the higher of the two.
That combination is frequently missed on forms, because each descriptor on its own looks like a minority case.
What the Rule Is Not
It is not an average of severity, and it is not a test of your worst day. GOV.UK frames the test as expecting "to have difficulty more than half of the days over a 12-month period", which is a count rather than a weighting. Both misreadings cost points, in opposite directions.
Averaging produces a middle description that matches no descriptor: "I can usually manage, with difficulty" is not a statement the assessment can score. Describing only crisis days invites the reply that the activity is possible on most days, which is the test that decides the point.
The accurate answer is a count. "I cannot do this on about five days a week, and on the other two I need an aid" gives a decision maker both the frequency and the descriptor.
Evidencing Frequency
A dated log over four to six weeks is the format the rule asks for. The rule compares days on which a descriptor applies with days on which it does not, and the DWP assessment guide treats variability as a matter for evidence rather than assumption. Nothing else demonstrates a count as directly.
Each entry needs three things: the date, the activity, and whether it was completed and how. Entries that record the days something did not happen are as useful as entries recording the days it did, because the rule is a comparison between the two.
Clinical letters rarely address frequency, because frequency is not what clinicians record. That is why a claim supported entirely by medical evidence can still fail this test.
How It Interacts With Reliability
The 50% rule decides which days count; the reliability criteria decide whether a day counts as done. They run in that order, and both must be satisfied.
An activity completed on a given day but unsafely, or in three times the usual time, or not repeatably, has not been completed to the descriptor's standard on that day. It therefore counts toward the days the descriptor applies rather than against them. The four tests are set out on the reliability criteria page.