Short answer
A lab thermometer is calibrated by comparing it with a reference thermometer that carries a traceable certificate, at the temperatures it is actually used at, and recording the error and correction at each point.
For the calibration of lab thermometers of the liquid-in-glass type, NIST adds a cheap in-house check between full calibrations: rest the thermometer 72 hours at room temperature, read it in an ice-point bath, and shift every certified correction by any change in that reading [1]. Whether the thermometer passes depends on the test it serves: a dissolution bath held at 37 ± 0.5 °C needs a far tighter thermometer than a disintegration bath held at 37 ± 2 °C [7][8].
What this revision corrects
The earlier version of this SOP gave a usable comparison method but no acceptance limit, no ice-point check and nothing for digital thermometers. Each change below is traced to its source.
| Earlier SOP said | Status | Verified position |
|---|---|---|
| Let thermometers stand 72 hours at room temperature | Kept, scoped | The 72 h rest is real, but it is the condition for an ice-point reading of a liquid-in-glass thermometer, because heating temporarily expands the bulb and depresses the ice point [1][2]. It does not apply to digital thermometers. |
| Heat oil in a beaker on a magnetic hotplate and read both thermometers | Replaced | An unstirred beaker on a hotplate has gradients between the two bulbs. Use a stirred liquid bath or a dry-block calibrator, with both bulbs side by side at the same depth. |
| No acceptance limit given | Added | Tolerance is set from the test the thermometer serves (section 8 and the limits table below). The thermometer can pass for one use and fail for another. |
| Calibrate at least twice a year, and new thermometers before use | Kept as site policy | No pharmacopoeia sets a thermometer interval. NIST recommends ice-point checks monthly at first, extended to every two months and then to once or twice a year if stable, and at least yearly even for thermometers used near room temperature [1][2]. WHO asks for a laboratory calibration plan [4]. |
| “Standard thermometer” not defined | Defined | The reference must be traceable; WHO says calibrations should be traceable to certified reference materials and SI units [4]. |
| Immersion depth mentioned without explanation | Expanded | A thermometer used at the wrong immersion can read with large errors; partial-immersion and total-immersion types are not interchangeable [1]. The stem correction is given in section 7. |
Swipe tables sideways on a phone.
The four stages at a glance
SOP for calibration of lab thermometers
The document below keeps the section order of the original SOP and can be adapted into your own QMS. Header fields are left blank on purpose.
1. Objective
To lay down the procedure for calibration of laboratory thermometers (liquid-in-glass and digital) used in the Quality Control laboratory, so that every temperature reading used in a test or a storage record carries a known correction.
2. Scope
Applies to all thermometers used for testing, water baths, dissolution and disintegration baths, incubators, refrigerators and sample storage in the QC department of [Company name]. It excludes fixed temperature sensors and data loggers of stability chambers, autoclaves and HVAC systems, which are calibrated under their own equipment SOPs.
3. Responsibility
- QC Chemist: inspection, ice-point checks, comparison calibration and records.
- Officer QC: review of records, calculation check and labelling.
- Head QA: approval of this SOP, tolerances and the calibration plan.
4. Accountability
Quality Control Manager.
5. Abbreviations
SOP: Standard Operating Procedure. QC: Quality Control. QA: Quality Assurance. LIG: liquid-in-glass. PRT: platinum resistance thermometer. NABL: National Accreditation Board for Testing and Calibration Laboratories.
6. Materials and equipment
- Reference thermometer with a valid calibration certificate from a calibration laboratory accredited to ISO/IEC 17025 (in India, NABL-accredited), covering the full working range.
- Dewar flask, shaved or crushed ice made from distilled or purified water, and distilled water.
- Stirred liquid bath (water up to about 80 °C; a suitable bath oil above that) or a dry-block calibrator with a well insert that fits both probes.
- Thermometer holder or clamp, lens or magnifier, lint-free tissue, calibration labels.
- Mercury spill kit wherever mercury-in-glass thermometers are still in use.
7. Procedure
7A. Pre-checks (all thermometers)
7B. Ice-point check (liquid-in-glass only)
7C. Comparison calibration (all thermometers)
error = test reading - true temperature and correction = -error.8. Acceptance criteria
The tolerance is not a property of the thermometer. It is derived from the temperature band of the test or storage condition the thermometer is assigned to. A widely used metrology convention is that the measuring device should be about four times more accurate than the band it checks; that ratio is a convention, not a pharmacopoeial requirement, so it is marked as site policy below. Each thermometer is labelled with the most demanding use it passes. See the limits table.
9. Frequency
- New thermometers: before first use (site policy, retained from the earlier SOP).
- Ice-point check of LIG thermometers: monthly at first, extended to every two months and then to once or twice a year if the ice point is stable; at least yearly even when used intermittently near room temperature [1][2].
- Full comparison calibration: at least twice a year (site policy), and after any repair, drop, overheating or out-of-range ice point.
- Digital thermometers: per the manufacturer’s recommendation or the laboratory’s own history, typically one year, shortened if records show drift [9].
10. Precautions
- Never use a thermometer past its due date or with its label missing; WHO expects calibration status to be identifiable on the equipment [4].
- Do not heat a LIG thermometer above its scale maximum; record any overheating.
- Handle mercury thermometers over a tray and keep the spill kit at hand. India ratified the Minamata Convention on mercury in June 2018; the convention’s phase-out date for mercury thermometers was 2020, and India sought an exemption until 2025 [10]. Plan replacement with mercury-free thermometers.
- Make a fresh ice bath each day it is needed [1].
11. Deviation handling
- If a thermometer fails its assigned tolerance, take it out of service and label it [4], or reassign it to a wider-tolerance use if it passes that.
- Raise a deviation and assess the effect on every test and storage record made with that thermometer since its last passing calibration.
- If the reference thermometer itself is found damaged or out of date, stop all calibrations until it is replaced.
12. Annexure-I: Thermometer calibration record
| Thermometer ID | Type / range | Immersion | Point (°C) | Reference reading | Ref. cert. correction | True temp. | Test reading | Error | Correction | Tolerance (use) | Result | Done by | Checked by |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| QC/TH/012 | LIG, -10 to 110 °C | Partial, 76 mm | 0.00 | – | – | 0.00 | +0.08 | +0.08 | -0.08 | Action limit (site) | Record shift | ||
| QC/TH/012 | LIG, -10 to 110 °C | Partial, 76 mm | 37 | 37.00 | +0.03 | 37.03 | 36.88 | -0.15 | +0.15 | ±0.5 (disintegration) | Pass | ||
Copies the table as tab-separated text that pastes into a spreadsheet. The filled rows are the worked examples below; delete them in your copy.
13. Annexure-II: Calibration status label
Thermometer ID · Calibrated on · Due on · Correction at use point(s) · Permitted use (for example “Disintegration, water bath; not for dissolution”) · Calibrated by · Checked by.
14. Revision history
| Version | Effective | Change |
|---|---|---|
| 00 | DD-MMM-YYYY | New SOP: comparison method only. |
| 01 | DD-MMM-YYYY | Added ice-point check, digital thermometers, use-based tolerances, stem correction, deviation handling and record format. |
Acceptance limits by use: where the tolerance comes from
The band in the second column is the requirement the thermometer helps to prove. The third column applies the four-to-one convention and is a site-policy suggestion only; your QA may set tighter or wider values.
| Use of the thermometer | Required band | Suggested thermometer tolerance | Basis |
|---|---|---|---|
| Dissolution medium (apparatus 1 to 4) | 37 ± 0.5 °C | ± 0.12 °C | compendial band Ph. Eur. 2.9.3 [7]; site policy tolerance |
| Disintegration immersion fluid | 37 ± 2 °C | ± 0.5 °C | compendial band USP 35 <701> [8]; site policy tolerance |
| Refrigerated sample storage | 2 to 8 °C | ± 0.75 °C | WHO QC laboratory practice cites refrigerated storage at 2 to 8 °C [4]; site policy tolerance |
| Ice-point shift between full calibrations | action limit | not more than the tightest tolerance assigned | site policy; method from NIST SP 1088 [1] |
| Melting range determinations | per monograph | per apparatus SOP | check source calibrated with reference substances; see the melting point SOP below |
| USP general chapter <21> Thermometers | not quoted | – | check source the current chapter text is behind the USP-NF paywall and could not be verified for this page |
Worked examples
Example 1: one thermometer, two verdicts
At 37 °C the reference reads 37.00 °C and its certificate correction is +0.03 °C, so the true temperature is 37.03 °C. The test thermometer reads 36.88 °C.
Against the ± 0.12 °C suggested for dissolution it fails. Against the ± 0.5 °C suggested for disintegration it passes. The label reads “Disintegration only”, and the deviation question is whether this thermometer was ever used to set a dissolution bath.
Example 2: an ice-point shift
The certificate gives the ice-point reading as 0.00 °C. Today, after a 72 h rest, the thermometer reads +0.08 °C in the ice bath. Following NIST’s method, every reading is now 0.08 °C higher than at certification, so 0.08 °C is subtracted from every certified correction [1]. A certified correction of +0.15 °C at 37 °C becomes +0.07 °C.
Example 3: why immersion matters
A total-immersion mercury thermometer reads 100 °C with 60 scale degrees of column standing above the bath at a mean 30 °C:
The emergent stem alone puts the reading two-thirds of a degree low [3]. At 37 °C the same effect is much smaller, which is why the error often goes unnoticed in routine use.
Liquid-in-glass vs digital lab thermometers
| Point | Liquid-in-glass | Digital (PRT, thermistor, thermocouple) |
|---|---|---|
| In-house check between calibrations | Ice-point check, after a 72 h rest [1] | Comparison at one use point against the reference |
| Can one point update the whole scale? | Yes for mercury up to 200 °C; not validated for organic liquids [1] | No; probe and electronics can drift differently across the range |
| How damage shows | Separated column, bubbles, foreign matter, cracks [1] | Often no visible sign; only a performance check shows it [9] |
| Immersion sensitivity | High; must match the marked immersion [1] | Lower, but the probe needs enough immersion to avoid stem conduction |
| Hazard | Mercury types: spill and disposal controls [5] | Battery and probe-cable damage |
| Standard for inspection | ASTM E77 covers visual inspection and scale accuracy verification of LIG thermometers [5] | Manufacturer’s specification and the calibration certificate |
For the Indian regulatory frame: the revised Schedule M notified by G.S.R. 922(E) of 28 December 2023 was written to harmonise Indian GMP with WHO guidelines [11], and WHO’s good practices for QC laboratories require a calibration plan, a schedule of due dates, equipment records and identification of calibration status [4]. Clause-level Schedule M text is not quoted on this page because the gazette copy could not be opened for this revision.
FAQ
No pharmacopoeia fixes the interval, so the laboratory sets it in its calibration plan. NIST suggests ice-point checks of liquid-in-glass thermometers monthly at first, extended to every two months and then once or twice a year if stable, and at least yearly. Many QC laboratories add a full comparison calibration twice a year and before first use.
Heating a liquid-in-glass thermometer expands its glass bulb, and the bulb takes time to contract back. Until it does, the ice-point reading is depressed. NIST specifies 72 hours at room temperature before the ice-point reading. The rule does not apply to digital thermometers.
For a mercury-in-glass thermometer used at 200 °C or below, yes: NIST treats a change at the ice point as a shift of the whole scale, so every certified correction moves by the same amount. NIST states that single-point recalibration has not been validated for organic-liquid thermometers, and it does not apply to digital ones, so those need a multi-point comparison.
It depends on the test it serves. A dissolution bath must hold 37 ± 0.5 °C, so a thermometer that is 0.15 °C out is too coarse for it, while the same thermometer is fine for a disintegration bath held at 37 ± 2 °C. Set the tolerance from the use, label the permitted use, and have QA approve the values.
When a total-immersion thermometer is used with part of its liquid column out of the bath, that part is cooler and the reading is low. The correction is k times n times the difference between bath and stem temperature, with k about 0.00016 per °C for mercury and about 0.001 per °C for organic liquids (NBS Monograph 150).
India ratified the Minamata Convention on mercury in June 2018. The convention set 2020 as the phase-out date for mercury thermometers, and UNEP reports that India sought an exemption until 2025. Check the current position for your site before buying replacements, and plan a move to mercury-free thermometers with their own calibration records.
Planning a new plant or expanding a QC laboratory?
Calibration SOPs are one part of a QC laboratory that has to pass inspection. Laafon’s plant setup consultancy works on facility design and setup for new and expanding pharmaceutical units across India.
Related on laafon.com
References
- Cross CD, Ripple DC, Strouse GF. Maintenance and validation of liquid-in-glass thermometers. NIST Special Publication 1088. Gaithersburg (MD): National Institute of Standards and Technology; 2009. Available from: https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication1088.pdf. Accessed October 2026.
- Wise JA. A procedure for the effective recalibration of liquid-in-glass thermometers. NIST Special Publication 819. Gaithersburg (MD): National Institute of Standards and Technology; 1991. Available from: https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication819.pdf. Accessed October 2026.
- Wise JA. Liquid-in-glass thermometry. NBS Monograph 150, section 5.3. Washington (DC): National Bureau of Standards; 1976. Available from: https://nvlpubs.nist.gov/nistpubs/Legacy/MONO/nbsmonograph150.pdf. Accessed October 2026.
- World Health Organization. WHO good practices for pharmaceutical quality control laboratories. Annex 1. WHO Technical Report Series No. 957. Geneva: WHO; 2010. Sections 7.10, 12 and 13. Available from: https://www.who.int/publications/m/item/trs957-annex1. Accessed October 2026.
- ASTM International. ASTM E77-14(2021) Standard test method for inspection and verification of thermometers. West Conshohocken (PA): ASTM International; 2021. Available from: https://store.astm.org/e0077-14r21.html. Accessed October 2026.
- ASTM International. ASTM E563-22 Standard practice for preparation and use of an ice-point bath as a reference temperature. West Conshohocken (PA): ASTM International; 2022. Available from: https://store.astm.org/e0563-22.html. Accessed October 2026.
- European Directorate for the Quality of Medicines. European Pharmacopoeia 7.0, 2.9.3 Dissolution test for solid dosage forms. Strasbourg: EDQM; 2010 (reprint edition; check the current edition). Available from: https://www.drugfuture.com/Pharmacopoeia/EP7/DATA/20903E.PDF. Accessed October 2026.
- United States Pharmacopeial Convention. USP 35-NF 30, general chapter 701 Disintegration. Rockville (MD): USP; 2012 (reprint edition; check the current edition). Available from: https://drugfuture.com/pharmacopoeia/usp35/data/v35300/usp35nf30s0_c701.html. Accessed October 2026.
- National Institute of Standards and Technology. Mercury thermometer alternatives: what is traceability? Gaithersburg (MD): NIST; 2011, updated 2024. Available from: https://www.nist.gov/pml/mercury-thermometer-alternatives/mercury-thermometer-alternatives-what-traceability. Accessed October 2026.
- UN Environment Programme, Global Mercury Partnership. India moves towards mercury-free healthcare sector. Nairobi: UNEP; 20 December 2024. Available from: https://www.unep.org/globalmercurypartnership/news/story/india-moves-towards-mercury-free-healthcare-sector. Accessed October 2026.
- Central Drugs Standard Control Organisation. Revised GMP guidelines and requirements under Schedule M [presentation]. New Delhi: CDSCO; 2024. Available from: https://www.pmda.go.jp/files/000269708.pdf. Accessed October 2026.
This procedure is a template for adaptation. It requires local qualification, validation and Quality Assurance approval before use, and every acceptance criterion must be verified against the current pharmacopoeial edition applicable at your site. Tolerances marked site policy are suggestions, not compendial requirements. Pharmacopoeial texts and Indian statutory instruments change between editions. Technical and educational content only, not medical, legal or investment advice. Reviewed by Darshan Singh, 23+ years in pharmaceutical QA, QC and regulatory affairs. Last updated 28 October 2026.
