What are the known risks or side effects of DMSO used by mouth or on skin?

The short version: the evidence here is about DMSO's local skin effects and its use as a drug carrier, not about systemic risks from oral or topical use in humans.

The retrieved studies tell you almost nothing about the risks of taking DMSO by mouth or putting it on your skin. One 2011 pooled safety analysis of 1,252 patients looked at a topical diclofenac solution that contained 45.5% DMSO [2]. The most common side effect was dry skin, which happened in 33% of the DMSO group versus 5% of the placebo group. Dyspepsia (upset stomach) occurred in 7.7% versus 2.9% [2]. Serious adverse events were rare and similar between groups. But that study tested a drug-plus-DMSO combination, not DMSO alone, so you cannot separate what the DMSO caused from what the diclofenac caused.

A 2002 case report describes a 19-year-old leukemia patient who got 99% DMSO applied topically for a chemotherapy extravasation [4]. He developed burning, swelling, and inflammation at the site. He died a month later from fungal infection, but that was from his underlying disease and chemotherapy, not from the DMSO.

The 2024 narrative review claims that high-purity DMSO (over 99.99%) reduces the garlic-like breath odor that was common with older, less pure DMSO preparations [5]. In one historical cohort using 90% DMSO, 5 of 6 inpatients and 9 of 10 outpatients reported halitosis. In a later cohort using 45.5% high-purity DMSO, only 12 of 1,243 patients reported it [5]. That is a real improvement, but the comparison is confounded by different doses and vehicles.

In the lab, DMSO is cytotoxic to human cells at concentrations above 1% [3]. At 5% and 10%, it kills cells at all time points. At 0.1% to 1%, it does not reduce viability but can actually increase mineralization activity [3]. These are in vitro findings, not human toxicity data, but they show DMSO is not biologically inert.

What is missing from this retrieval is striking. There are no long-term safety studies of oral DMSO in humans. No data on kidney or liver effects from repeated use. No studies comparing DMSO to placebo for any condition with hard clinical endpoints. The pharmacokinetic study in horses shows that oral DMSO is about 47% bioavailable and reaches peak blood levels within 45 minutes [1], but that tells you nothing about human safety.

My call: the evidence here is too thin to characterize the risks of DMSO taken by mouth or applied to skin. The known local effects are dry skin and irritation. The garlic breath is real but may be reduced with high-purity preparations. The systemic risks are unstudied in this retrieval. Confidence: low, because the retrieval contains no dedicated human safety trials of DMSO alone by either route.