What helps a cold sore: antivirals shorten it slightly, and only if started early

The most useful thing to know about a cold sore is that it heals on its own in about a week, and no treatment clears the underlying virus for good — so the honest goal is to shorten and soothe an episode, not cure it [s1]. Antiviral creams and pills do help, but modestly: the best trials show them trimming an episode by anywhere from about 18 hours to roughly a day [s1] [s3]. And they only work if started at the very first sign — the tingle or itch before a blister appears. Reach for one after the sore has erupted and much of the possible benefit is already gone.

A recurring virus, not a curable infection

Cold sores — recurrent herpes simplex labialis — affect an estimated 20% to 40% of the US population, and although the sores are self-limiting in healthy people, the discomfort and visibility send many to the pharmacy [s1]. The reason nothing cures them is biological: the herpes simplex virus stays latent in nerve tissue between outbreaks and reactivates periodically, so a treatment can shorten a flare but cannot evict the virus. This is a different virus from the one behind shingles, though both are herpesviruses, and the shingles vaccine does nothing for cold sores. It is also entirely distinct from canker sores — the mouth ulcers people often confuse with cold sores, which are not caused by a virus at all.

What the creams do

The over-the-counter option with real trial evidence is docosanol 10% cream. Two identical double-blind, placebo-controlled studies across 21 sites found that starting docosanol at the prodrome or early redness stage cut the median time to healing to 4.1 days in the 370 treated patients — about 18 hours shorter than in the 367 who got placebo (P = 0.008) — and shortened the time to the end of pain and other symptoms as well (P = 0.002) [s1]. It is a genuine effect, and a small one.

The prescription cream penciclovir tells the same story with different numbers. In a trial where 1,573 patients self-started treatment at the first symptom and applied the cream every two hours while awake for four days, healing of classical lesions was 0.7 day faster than with placebo — a median of 4.8 days versus 5.5 (hazard ratio 1.33, 95% confidence interval 1.18 to 1.49) — with pain resolving a little sooner too, at a median 3.5 versus 4.1 days [s2]. The chart shows how narrow that gap is: a demanding two-hourly regimen buys most people well under a day.

What the pills do

Oral antivirals are somewhat more effective than creams, mainly because they reach the virus throughout the body rather than sitting on the surface. Valaciclovir is better absorbed than older aciclovir, raising aciclovir levels three- to fivefold, which is the rationale for a short, high dose [s3]. In two randomized, placebo-controlled trials, taking 2 grams of valaciclovir twice daily for a single day, started at the first symptom of a cold sore, reduced the median duration of the episode by 1.0 day; a two-day regimen reduced it by 0.5 days [s3]. As with the creams, the benefit depends on starting at the first tingle — a patient-initiated approach, with the medicine already on hand, is what these trials tested [s3]. Prescription oral antivirals are drugs with dosing and safety considerations that belong to a clinician, not a self-treatment this article recommends.

The pattern across every option is the same one seen with antivirals for flu and COVID: the drugs are real, the effect is measured in hours to a day, and timing is everything. Much of what people credit to a dab of cream is also the episode simply running its course, the kind of improvement the placebo response in pain trials reliably produces.

What does not earn its billing

Worth naming plainly: there is no evidence that a cold sore can be aborted once a blister has formed, no treatment that prevents the virus from ever returning, and no cure. Lysine supplements, dabbed essential oils and similar remedies are widely sold but rest on far thinner evidence than the antivirals above, and none has trial support of the quality behind docosanol, penciclovir or valaciclovir [s1] [s2] [s3]. Keeping the area clean and using a lip barrier is reasonable comfort care; it is not treatment.

When a cold sore needs a doctor

The trials here studied ordinary recurrent cold sores in otherwise healthy people [s1]. Sores that are unusually severe or widespread, that will not heal, that recur very frequently, or that occur in someone with a weakened immune system or eczema fall outside that picture and warrant medical attention — as does any sore near the eye, which can threaten sight. Whether an antiviral, and which one, suits a given person is a clinical decision; what the evidence settles is that a cold sore was always going to heal on its own, and the most any remedy honestly offers is a slightly shorter, more comfortable week.