Water-based lube: how to choose— experience from 35 discussions and WHO criteria

Lubricants can be water-based, silicone-based or oil-based. Water-based ones are affordable and compatible with condoms and silicone toys, but you may need to reapply them during sex. Silicone ones last longer but can damage silicone toys. Oil-based ones stain fabric and are not suitable for latex condoms⁶.

Water-based lubricants differ a lot from one another, and the packaging barely shows it. Medically, the key measure is osmolality: how concentrated the solution is. Vaginal secretions normally have 260–290 mOsm/kg, semen 250–380, while most commercial lubricants have 2,000–6,000. Concentration is raised mainly by glycerin and propylene glycol¹.

Why it matters. In a laboratory model, mucosal irritation increased with osmolality: a gel at 316 mOsm/kg had no effect, while a gel at 5,848 caused severe irritation with tissue damage¹². On a three-dimensional model of the vaginal mucosa, all four lubricants above 1,500 damaged the epithelial layers, while there was no damage below 400¹³. Concentrated gels were toxic to cells, while those close to the body’s own secretions and silicone ones were not¹¹¹⁴. A caveat: these are data from laboratories and tissue samples; clinical studies are scarce.

“Glycerin-free” does not mean “low osmolality”. An annex to the WHO document lists Astroglide gel, free of glycerin and parabens, at 4,806 mOsm/kg — higher than KY Jelly at 2,007¹.

pH. Normal vaginal pH is 3.8–4.5, rectal pH about 7¹⁶. These are measures of the body, not requirements for a lubricant.

WHO standards. In the 2012 document, the WHO and UNFPA name an osmolality of no more than 380 mOsm/kg as ideal for a lubricant and 1,200 as an interim procurement limit: no more than 9.9% glycerin or 8.3% propylene glycol. For vaginal use they then advised a pH of about 4.5, for rectal use 5.5–7, and did not recommend above 7¹. In the 2026 specification, the osmolality limit remained below 1,200, and the pH range was widened to 4.0–7.0 with a tolerance of ±1.0. Preferred are lubricants that do not shift the pH of the vagina and rectum. They should contain no fragrances, colourings, spermicides, herbal or “pleasure-enhancing” additives. The recommendation to avoid polyquaternium was dropped as unsupported². The specification is written for procurement in reproductive health programmes, but its figures are also a clear benchmark for shoppers.

Clinical reviews advise choosing products as close as possible to natural secretions: products vary widely in pH and osmolality, and some may be harmful²⁰. A 2025 article in the Canadian journal for family physicians recommends water-based or silicone lubricants for people who use condoms or have recurrent infections and irritation — provided the lubricant is within the recommended osmolality and pH²¹.

What should not be in it. According to two Cochrane reviews, nonoxynol-9 does not protect against STIs or HIV, and genital lesions are more common with it⁹¹⁰; the CDC also advises avoiding it⁵. In the lab, the chlorhexidine in KY Jelly killed all three lactobacillus strains tested¹¹. On parabens, ISSM writes that their effect on vaginal flora and cells is debated and more research is needed⁶.

Glycerin and thrush. We found no human studies in which glycerin in a lubricant causes thrush. ISSM writes that glycerin and propylene glycol may, according to some data, damage tissue, but data are limited⁶. A Russian review from 2023, citing work on microbicides, says that 10% glycerin did not affect susceptibility to infection. Whether this was tested in humans is not clear from the review. The review was published with support from the company Laboratoriya Bioritm, and three of its six authors are company employees, although the authors declared no conflict of interest²². A link in humans was found for other practices. In a cohort of 141 women, oils inside the vagina were associated with Candida colonisation (44.4% versus 5%), and petroleum jelly with bacterial vaginosis (2.2 times higher risk)¹⁵. This is an association, not a proven cause.

Trying to conceive. In the lab, ordinary water-based gels reduce sperm motility by 60–100% within 60 minutes; canola oil, mineral oil and hydroxyethylcellulose-based gels do not⁷. But in cohorts of 296 and 6,467 women, lubricant did not reduce the chance of conception¹⁷¹⁸. A 2024 systematic review calls the data scarce¹⁹.

For dryness, the NHS distinguishes between a lubricant before sex and a vaginal moisturiser for ongoing dryness³. According to the Menopause Society’s position statement, most women with mild symptoms find over-the-counter products enough⁸.

If you have seen “pH 3.8–4.5” as a requirement for lubricant

It is easy to mistake this figure for a standard for lubricants. In fact, 3.8–4.5 is the pH of a healthy vagina¹⁶. For the lubricant itself, the WHO named about 4.5 in 2012 and allowed 4.0–7.0 in 2026¹². A Russian review from 2023 also gives 4–7 for lubricants²². Osmolality and an ingredient list without unnecessary additives matter more.

We analysed 44 discussions and personal stories on forums and in comments under articles. The unit of count is one thread or one story; several people often write within it: in total, more than 250 participants with personal experience. A topic was counted if at least one participant who had used a lubricant themselves wrote about it.

Of the 44, we excluded 9: seven threads with affiliate links to shops, one with a reply from a brand representative and an advertising review, and one from a clinic forum. All “X of 35” figures are based on the remaining 35 discussions. Platforms: Mumsnet (UK) — 20, Scarleteen (USA) — 7, Hellobee (USA) — 3, Patient.info (UK) — 2, Breast Cancer Now forum (UK) — 2, T—Zh (Russia) — 1. Languages: English — 34, Russian — 1. Collected on 11 November 2026; discussions from 2010–2026.

Limitations. Reddit, woman.ru and Yandex Reviews are closed to our tool, and we chose not to bypass their protection. There are no Ukrainian-language discussions and only one in Russian, so we draw no conclusions about Russian- and Ukrainian-speaking readers. Most threads come from a British forum, and 12 of 35 are about trying to get pregnant. The British brands Yes and Sylk, which are often praised (Yes in 7 of 35 discussions, Sylk in 4), are not sold at keks-shop. Coding was based on summaries of the pages, with a possible error of 1–2 discussions. Of the 24 keks-shop water-based lubricant pages we looked at, osmolality is listed on only one, and there are no independent measurements. Ingredients and product claims are taken from keks-shop pages on 11 November 2026.

People’s experience is not statistics: we show how many discussions mention something, not what percentage of people experience it. Medical facts come only from medical and scientific sources, each numbered in the list.