How Long to Take Antibiotics for Common Infections

For most common infections, the evidence increasingly supports shorter antibiotic courses than doctors have traditionally prescribed. Across multiple meta-analyses covering pneumonia, urinary tract infections, sinusitis, cellulitis, and ear infections, courses of about three to seven days perform as well as longer regimens in terms of symptom relief and cure rates, while causing fewer side effects. That said, the right duration still depends on which infection you have, its severity, and your overall health. Some infections genuinely require longer treatment, and cutting antibiotics short without guidance from your prescriber is not the same as your prescriber choosing a shorter course from the start.

Where the “Finish the Course” Rule Came From

The instruction to always finish your full course of antibiotics dates back to the earliest days of penicillin. Alexander Fleming warned about underdosing in his 1945 Nobel Prize speech, and the advice became medical gospel. The reasoning seemed intuitive: stop too soon, and surviving bacteria might develop resistance and bounce back. But that reasoning was never tested in rigorous clinical trials. The traditional course lengths for most infections were based on convention and caution rather than on data showing that a specific number of days was the minimum needed.

Recent evidence has flipped the concern. Evolutionary biology and clinical research now suggest that unnecessarily prolonged antibiotic exposure is itself a driver of resistance, because it gives bacteria more time under selective pressure to develop ways of surviving the drug.¹ When your immune system is expected to clear up the remaining bacteria on its own, shorter courses that get you past the worst of the infection may actually be the better strategy for both you and public health.² None of this means you should freelance your own duration. It means the prescribing world is catching up to the science, and your doctor may now give you a shorter prescription than you would have gotten a decade ago.

Urinary Tract Infections

Uncomplicated bladder infections in women are one of the clearest examples of shorter-is-fine. A Cochrane review comparing three-day antibiotic courses to five-to-ten-day courses found no meaningful difference in how quickly symptoms resolved, either in the short term or at longer follow-up. The longer courses did produce slightly better results on lab tests measuring whether bacteria were still detectable, but the difference in how patients actually felt was negligible. Meanwhile, side effects were significantly more common in the longer-treatment group.³

The picture gets more nuanced depending on which antibiotic is prescribed. A network meta-analysis in The Lancet Infectious Diseases found that for certain newer fluoroquinolones, even a single dose worked about as well as a three-day course. For co-trimoxazole and older quinolones, though, a single dose was clearly less effective than three days.⁴ The practical upshot: for a straightforward UTI, a three-day course of the right antibiotic is the standard in most guidelines. If your doctor prescribes seven or ten days for a simple bladder infection, it may be worth asking whether a shorter course would work. Complicated UTIs and kidney infections are a different story and typically still call for longer treatment.

Pneumonia

Community-acquired pneumonia is another infection where the evidence has shifted toward shorter treatment. A multicenter randomized trial found that stopping antibiotics based on clinical stability criteria after a minimum of five days produced outcomes no worse than the traditional seven-to-fourteen-day schedules for hospitalized patients.⁵ Multiple meta-analyses and randomized trials have confirmed that short courses of roughly three to seven days are not inferior to longer regimens for mild-to-moderate cases.⁶

Current international guidelines already reflect this. The key principle is clinical stability: once your fever has broken, your heart rate and breathing have settled, and you can eat and drink normally, most patients with non-severe pneumonia can safely stop antibiotics after about five days.⁷ Severe pneumonia, pneumonia caused by certain resistant organisms, or pneumonia in people with weakened immune systems may still warrant longer courses. But the old habit of prescribing ten to fourteen days for a garden-variety case no longer matches the data.

Sinus Infections

Most sinus infections are viral and do not need antibiotics at all. When a bacterial sinus infection is diagnosed and antibiotics are warranted, the question of how long to take them has a clear answer from a meta-analysis pooling twelve trials and over four thousand patients: short courses of three to seven days worked just as well as longer courses of six to ten days. Clinical success, bacterial clearance, relapse rates, and overall side effects were similar between groups. When the analysis zeroed in specifically on five-day versus ten-day regimens, the shorter course produced fewer adverse events.⁸

This is one of those infections where many patients end up taking antibiotics longer than necessary. Five days is enough for most uncomplicated cases. Chronic or recurrent sinusitis is different and warrants its own conversation with a specialist, but for a single episode of acute bacterial sinusitis, a short course does the job.

Strep Throat Is the Exception

Not every infection follows the shorter-is-better pattern. Strep throat caused by Group A streptococcus is a textbook case where the full ten-day course of penicillin still matters. The reason is specific: the goal is not just symptom relief but prevention of acute rheumatic fever, a serious inflammatory condition that can damage the heart. The American Heart Association’s scientific statement is unambiguous: penicillin for a full ten days (or a single injection of long-acting benzathine penicillin) remains the treatment of choice.⁹ A meta-analysis on the prevention of rheumatic fever confirmed that the infection is eradicated by ten days of oral penicillin or a single intramuscular injection.¹⁰

The reason strep throat stands apart from UTIs or sinusitis is that the antibiotic course is doing double duty: treating the sore throat and preventing a rare but devastating downstream complication. The evidence base for shorter courses simply is not there for this particular scenario, and the stakes of getting it wrong are high enough that the traditional duration remains justified.

Ear Infections in Young Children

Ear infections in kids are among the most common reasons antibiotics are prescribed in pediatrics, and the answer to “how long?” depends partly on the child’s age. For children two and older, the evidence is encouraging for shorter courses. A large network meta-analysis of 89 randomized trials found that seven-day regimens of amoxicillin or amoxicillin-clavulanate were not inferior to ten-day courses for clinical response in children over two.¹¹

For younger children, especially those under two, the picture is different. A New England Journal of Medicine trial found that five days of amoxicillin-clavulanate led to significantly more treatment failures than ten days in children six months to two years old. About a third of children on the short course had clinical failure, compared with roughly one in six on the full course.¹² Despite this evidence, prescribing patterns have been slow to differentiate by age. A study across two large health systems found that three-quarters of all ear infection prescriptions for children two and older were still for ten days, with only a fifth for seven days and a mere five percent for five days.¹³ That suggests a gap between the evidence and routine practice.

Skin Infections

Uncomplicated cellulitis, the red, swollen, warm patches of infected skin that send people to urgent care, responds well to a short course. A controlled comparison of five-day versus ten-day treatment found identical cure rates: 98% in both groups at both two-week and four-week follow-up.¹⁴ For straightforward cellulitis without abscesses or deep tissue involvement, five days appears sufficient. Complicated skin infections involving deeper structures, abscesses that need drainage, or infections in people with diabetes or poor circulation may need longer treatment and closer follow-up.

Why Longer Courses Can Actually Be Harmful

Beyond the abstract concern about antibiotic resistance at the population level, there are concrete personal risks to taking antibiotics longer than necessary. A longitudinal study of over 100,000 patients found that longer antibiotic courses were associated with a meaningfully higher risk of Clostridioides difficile infection, a potentially dangerous gut infection. Compared to a seven-day course, a ten-day course carried about 12% more risk, and a fourteen-day course about 27% more risk.¹⁵

The effect on your gut microbiome is also dose-dependent. In a controlled mouse study, three days of amoxicillin caused modest disruption to gut bacteria that recovered relatively quickly, while seven- and fourteen-day courses caused dramatic drops in bacterial richness and diversity. The fourteen-day group’s microbiome took about a week longer to bounce back than the seven-day group’s.¹⁶ Human data from a trial nested within a randomized study showed that both seven- and fourteen-day antibiotic groups had reduced gut diversity compared to untreated controls, though the two antibiotic groups looked similar to each other through day fourteen.¹⁷ The animal data suggests meaningful differences between short and long courses, while in critically ill humans the picture is murkier, possibly because other factors in hospital settings dominate.

The Big-Picture Safety Data

A systematic review and meta-analysis comparing shorter versus longer courses across multiple types of infections in hospitalized patients found no significant differences in clinical cure rates, bacterial clearance, short-term death, longer-term death, or infection recurrence. The differences between groups were consistently small and statistically indistinguishable from zero.¹⁸ When combined with the side-effect and resistance advantages of shorter courses, this body of evidence explains why stewardship programs at hospitals worldwide have been pushing prescribers toward the minimum effective duration.

In the UK, adoption of national guidelines recommending shorter courses for common infections was estimated to reduce total antibiotic use in emergency departments by about 7.5%.¹⁹ That may not sound dramatic, but scaled across an entire health system, it represents a meaningful reduction in resistance pressure and unnecessary side effects.

When You Should Not Expect a Shorter Course

People with weakened immune systems, including organ transplant recipients, people on chemotherapy, and those with conditions like HIV, have historically been left out of the clinical trials comparing short and long antibiotic courses. A review in Clinical Microbiology and Infection noted that while retrospective studies in transplant recipients have shown similar outcomes with shorter courses, and some data in cancer patients with low white blood cell counts look promising, the evidence base is thinner and the stakes are higher.²⁰ Guidelines themselves often fail to address these populations directly.²¹

Other situations that typically call for longer or carefully monitored courses include:

• Bone and joint infections: these often require weeks of treatment because antibiotics penetrate bone poorly.
• Heart valve infections (endocarditis): four to six weeks is standard because the infection sits in an area with limited blood flow to deliver the drug.
• Tuberculosis: six months or more, due to the slow-growing nature of the bacteria.
• Complicated or deep abscesses: may need longer courses in addition to drainage.

The short-course evidence applies primarily to common, uncomplicated outpatient infections. If you have been prescribed a longer course for something more serious, that prescription likely reflects a genuine need.

Why People Stop Early and Why That Is a Separate Problem

Even when a doctor prescribes what they believe is the right duration, a large fraction of patients do not finish. A prospective study from Eastern Ethiopia found that about 35% of outpatients stopped their oral antibiotics before the prescribed end date. The most powerful predictor was course length: prescriptions of seven days or longer were nearly three times as likely to be abandoned as shorter ones.²² A randomized study found that pharmacist-led education significantly improved adherence, and that the most common reasons people gave for stopping early were feeling better and simply forgetting to take doses.²³

This is where the conversation gets tricky. On one hand, the emerging evidence says many prescribed courses are longer than they need to be, so some patients who stop early may coincidentally land near the right duration. On the other hand, unilaterally deciding to stop on day three of a ten-day strep throat prescription because you feel fine is not the same as your doctor choosing a five-day course based on clinical data. The former is a gamble; the latter is evidence-based practice. The distinction matters. If you feel your course seems long, ask your prescriber whether a shorter duration is appropriate rather than quietly skipping pills.

Blood Tests That Could Personalize Duration

One of the most promising developments in antibiotic stewardship is the use of blood markers to guide when to stop treatment. Procalcitonin, a protein that rises in response to bacterial infection and falls as the infection resolves, has been tested as a stop signal for antibiotics. A large randomized trial in critically ill patients with suspected sepsis found that using daily procalcitonin measurements to guide decisions reduced antibiotic duration by roughly a day on average compared to standard care, without increasing death rates.²⁴ An earlier trial reported that procalcitonin guidance not only cut treatment duration but was associated with lower mortality, possibly because it helped doctors distinguish true bacterial infections from other causes of illness sooner.²⁵

For outpatient infections, the tools are less developed. The appealing idea of simply telling patients “stop when you feel better” has not yet been validated in primary care settings. A 2024 review noted that no studies have established the safety of symptom-guided discontinuation for common outpatient infections, and significant challenges remain before that approach could be recommended broadly.²⁶ Until that research arrives, the safest approach remains following the duration your prescriber gives you and having a conversation if you have concerns about it.

Dental Infections and Surgical Prophylaxis

Dental prescribing is an underappreciated contributor to total antibiotic use. Dentists often prescribe antibiotics before or after extractions and implant procedures, sometimes for days longer than necessary. Current dental guidance supports short courses and clinical reassessment but offers limited structure for deciding when to stop once treatment has started.²⁷ For procedures like extractions and implants, evidence suggests that a single pre-procedure dose may be all that is needed, and switching to that model could substantially reduce overprescribing in dentistry.²⁸

If your dentist hands you a prescription for five or seven days of antibiotics after a routine extraction, it is reasonable to ask whether a shorter course or a single preventive dose would suffice. Dental infections that have already established themselves, like a spreading abscess with facial swelling, are a different matter and may need a full therapeutic course. But prophylactic prescribing around procedures is one of the areas with the widest gap between what the evidence supports and what routinely happens.