The sections below cover how large that swing is, why it happens, how much the number also moves from one test to the next, and why a borderline result is normally repeated rather than acted on straight away.
TSH, or thyroid stimulating hormone, is the pituitary gland's signal to the thyroid, reported in blood as mIU/L. The pituitary senses how much thyroid hormone is circulating and adjusts its signal accordingly, so TSH usually moves in the opposite direction to thyroid hormone output. When circulating thyroid hormone runs low, the pituitary pushes harder and TSH rises. When it runs high, the signal quiets and TSH falls. That inverse relationship is why TSH is often the first thyroid number a clinician looks at. The full detail on what TSH is, how the test is run, what the reference ranges mean and how a high or low result is read sits on the biomarker page. This guide stays on one question, how far the number moves within a day and between draws.
The gap is big enough to matter at the edges of a reference range. Research suggests that serum TSH varies within the same person over hours, days and seasons, and that the circadian rhythm is one of the main sources of that within-person variation 1. Research suggests that, in healthy adults, the night-time peak runs roughly double the late afternoon low 23.
Put that in practical terms. If an afternoon sample lands comfortably in the middle of a typical reference range, a sample drawn from the same person in the small hours could sit near the upper end of that same range, with nothing different happening in the thyroid gland itself. The thyroid has not changed output. The pituitary is simply at a different point in its daily cycle.
Most routine draws never capture the full span, because blood is usually taken during clinic hours rather than at two in the morning. Within a normal working day the difference between an early appointment and a late one is smaller than the full peak to trough distance, but it is still present, because the morning hours sit on the falling limb between the overnight peak and the afternoon low 23.
TSH is lowest in the late afternoon and early evening and highest around the middle of the night. Research suggests that levels begin climbing in the evening, reach a peak in the hours around midnight and the early morning, then decline through the morning toward an afternoon trough 23.
The table below summarises that shape. It describes the general pattern reported in healthy adults on a conventional sleep schedule, not a timetable that every individual follows exactly.
| Time window | What TSH is generally doing |
|---|---|
| Mid to late afternoon | At or near its daily low point 23 |
| Evening | Rising from the afternoon trough 2 |
| Around midnight to early morning hours | At or near its daily peak 23 |
| Morning clinic hours | Falling from the overnight peak, still above the afternoon low 23 |
Two things follow from this. First, an 8 am draw and a 4 pm draw are not interchangeable samples, even on the same day in the same person. Second, the rhythm is anchored to the sleep and wake cycle, so someone who sleeps during the day will not necessarily match the pattern in the table. For anyone tracking a number over months, the useful habit is to hold the hour of the draw roughly constant, so that changes in the result reflect physiology rather than scheduling.
The night-time rise comes from the brain, not the thyroid. The hypothalamus and pituitary run on a circadian timing system that is coupled to the sleep and wake cycle, and TSH release follows that timing, which is why the signal climbs in the evening and peaks overnight before easing through the day 12.
There is a second layer on top of the clock. Sleep itself appears to restrain the nocturnal surge, so the rhythm reflects both the internal timer and what the body is actually doing at that hour. Research suggests that sleep and circadian factors are among the biological mechanisms behind within-person variation in TSH, alongside seasonal change, acute illness and differences between laboratory assays 1.
The thyroid gland responds slowly compared with the pituitary signal. Circulating thyroid hormone is buffered by binding proteins and a long turnover time, so the overnight TSH surge does not translate into a matching overnight swing in thyroid hormone levels. That mismatch is useful to understand, because it explains why TSH is the thyroid number most sensitive to the hour of the draw while free T4 is comparatively steady. When a result moves between tests and the thyroid hormone values have barely budged, timing and normal within-person variation are among the first explanations a clinician considers, rather than a change in gland function.
It can, particularly for values sitting near a boundary. Research suggests that in people with hypothyroidism, TSH measured in the afternoon is lower than in an early morning sample from the same person, by enough that the time of the draw can change whether a result falls inside or outside the reference range 4. The biology is the same in people without a thyroid diagnosis; the difference is that a borderline value is where the timing effect changes the interpretation.
| Marker | Typical lab reference range | FitScript optimal range |
|---|---|---|
| TSH (serum) | 0.45 to 4.5 mIU/L | 0.5 to 2 mIU/L |
Reference ranges vary by laboratory, assay, age and sex, so the figures on your own report are the ones that apply to your sample. The FitScript optimal range is an optimization target, not a diagnostic cut-off, and a value outside it is not a diagnosis of anything.
The practical consequence is worth stating plainly. A result of 4.2 mIU/L from an early morning draw and a result of 3.4 mIU/L from a late afternoon draw may describe the same underlying setpoint seen at two points on the same curve. Neither number on its own establishes what is happening. What matters is the pattern across repeated samples taken under comparable conditions, read next to symptoms, thyroid hormone levels and any antibody results by a clinician who can see the whole picture.
It varies more than most people expect, and the daily rhythm is only part of it. Research suggests that an individual's TSH fluctuates around a personal setpoint over the long term, with the variation within one person narrower than the spread seen across the population as a whole 5. That is the key idea behind repeat testing. Your own baseline is a tighter target than the wide reference range printed on the report.
Because the individual range is narrower than the population range, a move that looks small against the reference interval can still be a real move for you. Someone who has sat at 1.2 mIU/L on four occasions and then returns 2.8 mIU/L is still inside a typical reference range, yet the change is large relative to their own history. Equally, a single reading that sits slightly outside the range may be an ordinary excursion rather than a shift in thyroid function.
Several sources stack on top of each other across repeat draws. Research suggests that within-person variation in TSH arises from the circadian rhythm, day to day biological variation, seasonal effects, intercurrent illness and analytical differences between assays 1. None of these are errors. They are the normal behaviour of a signal that is constantly adjusting. Reading a single value as a fixed property of your thyroid overstates what one blood sample can tell you, which is the practical argument for serial measurement under consistent conditions.
Because one borderline value cannot distinguish a genuine shift from ordinary variation. Research suggests that a single TSH measurement may not represent an individual's usual setpoint, given how much the number moves within a person over time 15. A repeat draw, taken at a comparable hour under comparable conditions, separates a one off reading from a trend.
This is standard clinical caution rather than scepticism about the lab. Consider the sequence. The observation is a value a little above or below the reference range. The possible meanings include a true change in thyroid signalling, a draw taken at a high point of the daily curve, a recent illness, or simple biological variation. The next step is not a conclusion. It is a second measurement, usually alongside thyroid hormone levels and sometimes thyroid antibodies, so the clinician has more than one data point to reason from.
The timing of the repeat matters as much as the repeat itself. Comparing a morning value with an afternoon value introduces exactly the variation you are trying to see past 4. Comparing two samples drawn at a similar hour removes most of that noise.
Nothing here argues for ignoring an unusual result. An abnormal value is information and belongs in a conversation with a clinician. It simply does not, on its own, settle the question. Any decision about what follows, including whether treatment of any kind is appropriate, belongs to a licensed provider who has reviewed the person as well as the numbers.
Several things besides the hour of the draw. Research suggests that the main sources of within-person variation in serum TSH include circadian rhythm, day to day biological variation, seasonal change, acute illness and differences between laboratory assays 1. The table groups the practical ones.
| Factor | Why it can shift the number |
|---|---|
| Time of day | The pituitary runs TSH on a daily cycle, high overnight and low in the afternoon 23 |
| Day to day variation | TSH moves around a personal setpoint rather than holding a fixed value 5 |
| Season | Seasonal change is one of the recognised sources of within-person variation 1 |
| Acute illness | Short term illness is a recognised source of variation in TSH 1 |
| Sleep and circadian disruption | Sleep and circadian factors sit among the mechanisms behind TSH variability 1 |
| Laboratory and assay | Different platforms do not always return identical values, so results from two labs may not line up exactly 1 |
There are also medical situations, including pregnancy and the use of certain prescribed medicines, where expected TSH behaviour differs. Those belong in a conversation with a clinician who knows your history, because the interpretation changes with the context rather than following a general rule.
It can, because the TSH rhythm is tied to the sleep and wake cycle rather than to the clock on the wall. Research suggests that sleep and circadian mechanisms underlie part of the within-person variation seen in serum TSH 1, and the daily curve itself is described in relation to the normal sleep schedule of the people studied 2.
For someone who works nights or rotates shifts, the implication is straightforward. The relationship between the hour on the clock and the point on their TSH curve is not the same as it is for a day worker, so a 9 am draw after a night shift is not equivalent to a 9 am draw after a full night of sleep. That does not make the result wrong. It makes the context essential when the result is read.
The same logic applies to one off disruptions such as an overnight flight, a broken night with a sick child, or a late finish before an early appointment. These are not reasons to postpone testing indefinitely, and they are certainly not reasons to dismiss a result that comes back outside the range. They are details worth noting on the day so that the person interpreting the result knows what the sample reflects.
If you track TSH over time, the most useful thing you control is consistency. Same approximate hour, similar sleep the night before, similar fasting status. Consistency does not remove biological variation, but it strips out the part of the noise that comes from scheduling.
The nocturnal rise is not fixed and can be altered in some clinical states. Research suggests that the circadian rhythm of plasma TSH can be blunted in people with depression and that the pattern may change with clinical recovery 6.
This is useful context rather than a test you would order for that purpose. It illustrates that the shape of the daily TSH curve reflects central regulation, and that central regulation responds to more than thyroid hormone levels alone. A single daytime TSH drawn in the middle of this picture tells you where the signal was at that hour. It does not describe the rhythm behind it.
For a reader looking at their own results, the takeaway is modest and practical. Context shapes interpretation. A number that looks unexpected deserves a conversation about what else was happening around the time of the draw, including sleep, illness, stress and mood, before anyone reads significance into it.
Consistency matters more than finding one perfect hour. Most laboratory reference ranges are built from samples collected during normal working hours, and most clinical practice follows a morning draw, so a morning appointment keeps your result comparable both to the reference range and to your own prior tests.
The argument for morning sampling is that it is the conventional comparison point and easy to repeat. The argument against treating it as a rule is that morning sits on the falling limb from the overnight peak, so morning values tend to run higher than afternoon values from the same person 23, and in people with hypothyroidism that difference can be enough to change which side of the reference boundary a result lands on 4.
So the sensible approach is not to chase a specific hour but to stop changing it. Pick a window you can repeat, book follow up draws into the same window, and record the actual time on each occasion. If a result sits near a decision boundary, that recorded time becomes genuinely useful information for the clinician reviewing it.
Fasting status deserves the same treatment. Whether or not your lab asks you to fast, keeping the same approach each time removes one more variable from the comparison. Follow whatever instruction your laboratory gives for the specific panel being drawn.
Four details carry most of the value, and all of them take seconds to record. The time of the draw, whether you had eaten, how you slept the night before, and whether you were unwell or recently unwell in the preceding weeks. Research suggests that acute illness and circadian factors are among the recognised sources of within-person variation in TSH 1, which is exactly why these notes earn their place.
It is also worth noting which laboratory ran the sample. Results from different platforms do not always align perfectly, so a shift between two numbers can partly reflect the change of lab rather than a change in you 1. Where you can, keep testing with the same provider.
Add any recent changes in routine that affect sleep, such as a run of night shifts, long haul travel or a period of broken sleep. Add anything your clinician has prescribed or changed, without trying to judge its effect yourself, since that assessment belongs to the prescriber.
Kept together, these notes convert a scattered set of numbers into a readable series. A clinician looking at four TSH values with times and conditions attached can separate the pattern from the noise. The same four values with no context invite over-reading of differences that were never biological in the first place.
TSH on its own describes the signal, not the output, so it is usually read next to the hormones it is responding to. Free T4 measures the main circulating thyroid hormone the pituitary is reacting to, and free T3 reflects the more active form that tissues use. When TSH moves and free T4 holds steady, the interpretation is different from when both move together, which is one reason a single number is rarely the end of the conversation.
Thyroid antibodies, particularly thyroid peroxidase antibodies, indicate whether the immune system is involved in what the thyroid is doing. They answer a different question from TSH, and they are typically ordered when a pattern of results prompts it rather than routinely.
It is also common to look at markers that sit around thyroid physiology rather than inside it, since nutritional status, inflammation and metabolic markers all form part of the context a clinician weighs. A labs-first membership such as FitScript tests 150+ biomarkers per year and has Atlas, its AI health advisor, interpret them in combination rather than one line at a time, which is the level at which a thyroid picture actually becomes readable.
Which of these are appropriate for you is a clinical decision. A licensed provider decides what to order, how to sequence it and what any combination of results means in your particular case.
Any result outside your laboratory's reference range is worth discussing, and so is a result inside the range that has moved noticeably against your own history. The reasoning follows the same three steps each time. The observation is the number. The possible meanings include a true shift in thyroid signalling, the hour of the draw, recent illness or ordinary within-person variation 1. The next step is a conversation, usually with a repeat test, not a conclusion.
Bring symptoms into that conversation, because they change how a borderline number is weighed. Persistent fatigue, unexplained weight change, cold or heat intolerance, hair or skin changes, low mood, palpitations, or changes in menstrual pattern are all relevant history. So is pregnancy, planning a pregnancy, a family history of thyroid disease, or a previous thyroid diagnosis.
A markedly abnormal value, or a result accompanied by significant symptoms, is worth discussing with a clinician promptly rather than at the next routine review. A value that is only slightly suboptimal is still worth raising, simply with less urgency. Never set an abnormal result aside because the draw was at an unusual hour; the timing is context for the clinician, not a reason to dismiss the number.
This article is educational and is not medical advice. Discuss your own results, and any treatment, with a clinician or licensed provider who can evaluate your full history.
For the habits and follow up testing that sit behind a number trending upward over time, see the guide on how to lower TSH naturally.
The number is generally higher in the early morning and lower in the mid to late afternoon, because the pituitary runs TSH on a daily cycle that peaks overnight. The gap can be wide enough to move a borderline value across a reference boundary, which is why the time of the draw is worth noting on every test.
It matters less. A value sitting comfortably in the middle of your lab's reference range is unlikely to be reclassified by the hour of the draw. Timing matters most when a result sits close to the upper or lower boundary, or when you are comparing one test with another to look for a trend.
Follow whatever instruction your laboratory gives for the panel being drawn, since requirements differ by lab and by what else is being measured in the same sample. Whichever approach applies, keeping it the same each time makes your results easier to compare across tests.
Because a single value near a boundary cannot separate a real change in thyroid signalling from the normal movement of the number across a day and between days. A second sample, drawn at a similar hour under similar conditions, gives the clinician two data points instead of one before any decision is made.
No. A single value out of range is information, not a diagnosis. It may reflect the hour of the draw, a recent illness or ordinary variation, or it may reflect a genuine change. A clinician works that out using repeat testing, thyroid hormone levels, antibodies where appropriate, and your symptoms and history.
Not necessarily better, but usually more practical. Morning is the conventional sampling window, so it keeps your result comparable with reference ranges and with your own previous tests. The more important habit is picking a time window you can repeat and sticking to it.
It can, because the TSH rhythm is tied to the sleep and wake cycle rather than to the clock time of your appointment. That is not a reason to postpone testing or to dismiss a result. Note the disruption on the day so that whoever reviews the result knows what the sample reflects.
That interval is a clinical decision and depends on how far the value sits from the range, your symptoms and your history. Ask the clinician who ordered the test when to repeat it, and book the repeat into the same time window as the original draw so the two values are comparable.
Free T4 and free T3 are commonly measured with it, because TSH describes the signal while those describe the hormone output the signal is responding to. Thyroid antibodies are sometimes added depending on the pattern of results. A licensed provider decides what to order and how to read the combination.