Antibiotics improve outcomes in those with bacterial pneumonia. Antibiotic choice depends initially on the characteristics of the person affected, such as age, underlying health, and the location the infection was acquired. In the UK, treatment before culture results with amoxicillin is recommended as the first line for community-acquired pneumonia, with doxycycline or clarithromycin as alternatives. In North America, where the "atypical" forms of community-acquired pneumonia are more common, macrolides (such as azithromycin or erythromycin), and doxycycline have displaced amoxicillin as first-line outpatient treatment in adults. In children with mild or moderate symptoms, amoxicillin remains the first line. The use of fluoroquinolones in uncomplicated cases is discouraged due to concerns about side-effects and generating resistance in light of there being no greater clinical benefit.

For those who require hospitalization and caught their pneumonia in the community the use of a β-lactam such as cephazolin plus macrolide such as azithromycin or a fluoroquinolones is recommended. The addition of corticosteroids also appears to improve outcomes.

The duration of treatment has traditionally been seven to ten days, but increasing evidence suggests that shorter courses (three to five days) are similarly effective. Recommendations for hospital-acquired pneumonia include third- and fourth-generation cephalosporins, carbapenems, fluoroquinolones, aminoglycosides, and vancomycin. These antibiotics are often given intravenously and used in combination. In those treated in hospital, more than 90% improve with the initial antibiotics.

Antibiotics are the treatment of choice for bacterial pneumonia, with ventilation (oxygen supplement) as supportive therapy. The antibiotic choice depends on the nature of the pneumonia, the microorganisms most commonly causing pneumonia in the geographical region, and the immune status and underlying health of the individual. In the United Kingdom, amoxicillin is used as first-line therapy in the vast majority of patients acquiring pneumonia in the community, sometimes with added clarithromycin. In North America, where the "atypical" forms of community-acquired pneumonia are becoming more common, clarithromycin, azithromycin, or fluoroquinolones as single therapy have displaced the amoxicillin as first-line therapy.

Local patterns of antibiotic-resistance always need to be considered when initiating pharmacotherapy. In hospitalized individuals or those with immune deficiencies, local guidelines determine the selection of antibiotics.

Neuraminidase inhibitors may be used to treat viral pneumonia caused by influenza viruses (influenza A and influenza B). No specific antiviral medications are recommended for other types of community acquired viral pneumonias including SARS coronavirus, adenovirus, hantavirus, and parainfluenza virus. Influenza A may be treated with rimantadine or amantadine, while influenza A or B may be treated with oseltamivir, zanamivir or peramivir. These are of most benefit if they are started within 48 hours of the onset of symptoms. Many strains of H5N1 influenza A, also known as avian influenza or "bird flu", have shown resistance to rimantadine and amantadine. The use of antibiotics in viral pneumonia is recommended by some experts, as it is impossible to rule out a complicating bacterial infection. The British Thoracic Society recommends that antibiotics be withheld in those with mild disease. The use of corticosteroids is controversial.

Treatment of CAP in children depends on the child's age and the severity of illness. Children under five are not usually treated for atypical bacteria. If hospitalization is not required, a seven-day course of amoxicillin is often prescribed, with co-trimaxazole an alternative when there is allergy to penicillins. Further studies are needed to confirm the efficacy of newer antibiotics. With the increase in drug-resistant Streptococcus pneumoniae, antibiotics such as cefpodoxime may become more popular. Hospitalized children receive intravenous ampicillin, ceftriaxone or cefotaxime, and a recent study found that a three-day course of antibiotics seems sufficient for most mild-to-moderate CAP in children.

"Streptococcus pneumoniae" — amoxicillin (or erythromycin in patients allergic to penicillin); cefuroxime and erythromycin in severe cases.

"Staphylococcus aureus" — flucloxacillin (to counteract the organism's β-lactamase).

Most newborn infants with CAP are hospitalized, receiving IV ampicillin and gentamicin for at least ten days to treat the common causative agents "streptococcus agalactiae", "listeria monocytogenes" and "escherichia coli". To treat the herpes simplex virus, IV acyclovir is administered for 21 days.

Usually initial therapy is empirical. If sufficient reason to suspect influenza, one might consider oseltamivir. In case of legionellosis, erythromycin or fluoroquinolone.

A third generation cephalosporin (ceftazidime) + carbapenems (imipenem) + beta lactam & beta lactamase inhibitors (piperacillin/tazobactam)

While antibiotics with activity specifically against "M. pneumoniae" are often used (e.g., erythromycin, doxycycline), it is unclear if these result in greater benefit than using antibiotics without specific activity against this organism in those with an infection acquired in the community.

Treatment of VAP should be matched to known causative bacteria. However, when VAP is first suspected, the bacteria causing infection is typically not known and broad-spectrum antibiotics are given (empiric therapy) until the particular bacterium and its sensitivities are determined. Empiric antibiotics should take into account both the risk factors a particular individual has for resistant bacteria as well as the local prevalence of resistant microorganisms. If a person has previously had episodes of pneumonia, information may be available about prior causative bacteria. The choice of initial therapy is therefore entirely dependent on knowledge of local flora and will vary from hospital to hospital. Treatment of VAP with a single antibiotic has been reported to result in similar outcomes as with a combination of more than one antibiotics, in terms of cure rates, duration of ICU stay, mortality and adverse effects.

Risk factors for infection with an MDR strain include ventilation for more than five days, recent hospitalization (last 90 days), residence in a nursing home, treatment in a hemodialysis clinic, and prior antibiotic use (last 90 days).

Possible empirical therapy combinations include (but are not limited to):

- vancomycin/linezolid and ciprofloxacin,

- cefepime and gentamicin/amikacin/tobramycin

- vancomycin/linezolid and ceftazidime

- Ureidopenicillin plus β-lactamase inhibitor such as piperacillin/tazobactam or ticarcillin/clavulanate

- a carbapenem (e.g., imipenem or meropenem)

Therapy is typically changed once the causative bacteria are known and continued until symptoms resolve (often 7 to 14 days). For patients with VAP not caused by nonfermenting Gram-negative bacilli (like Acinetobacter, Pseudomonas aeruginosa) the available evidence seems to support the use of short-course antimicrobial treatments (< or =10 days).

People who do not have risk factors for MDR organisms may be treated differently depending on local knowledge of prevalent bacteria. Appropriate antibiotics may include ceftriaxone, ciprofloxacin, levofloxacin, or ampicillin/sulbactam.

As of 2005, there is ongoing research into inhaled antibiotics as an adjunct to conventional therapy. Tobramycin and polymyxin B are commonly used in certain centres but there is no clinical evidence to support their use.

Patients with HCAP are more likely than those with community-acquired pneumonia to receive inappropriate antibiotics that do not target the bacteria causing their disease.

In 2002, an expert panel made recommendations about the evaluation and treatment of probable nursing home-acquired pneumonia. They defined probably pneumonia, emphasized expedite antibiotic treatment (which is known to improve survival) and drafted criteria for the hospitalization of willing patients.

For initial treatment in the nursing home, a fluoroquinolone antibiotic suitable for respiratory infections (moxifloxacin, for example), or amoxicillin with clavulanic acid plus a macrolide has been suggested. In a hospital setting, injected (parenteral) fluoroquinolones or a second- or third-generation cephalosporin plus a macrolide could be used. Other factors that need to be taken into account are recent antibiotic therapy (because of possible resistance caused by recent exposure), known carrier state or risk factors for resistant organisms (for example, known carrier of MRSA or presence of bronchiectasis predisposing to Pseudomonas aeruginosa), or suspicion of possible Legionella pneumophila infection (legionnaires disease).

In 2005, the American Thoracic Society and Infectious Diseases Society of America have published guidelines suggesting antibiotics specifically for HCAP. The guidelines recommend combination therapy with an agent from each of the following groups to cover for both "Pseudomonas aeruginosa" and MRSA. This is based on studies using sputum samples and intensive care patients, in whom these bacteria were commonly found.

- cefepime, ceftazidime, imipenem, meropenem or piperacillin–tazobactam; plus

- ciprofloxacin, levofloxacin, amikacin, gentamicin, or tobramycin; plus

- linezolid or vancomycin

In one observational study, empirical antibiotic treatment that was not according to international treatment guidelines was an independent predictor of worse outcome among HCAP patients.

Guidelines from Canada suggest that HCAP can be treated like community-acquired pneumonia with antibiotics targeting Streptococcus pneumoniae, based on studies using blood cultures in different settings which have not found high rates of MRSA or Pseudomonas.

Besides prompt antibiotic treatment, supportive measure for organ failure (such as cardiac decompensation) are also important. Another consideration goes to hospital referral; although more severe pneumonia requires admission to an acute care facility, this also predisposes to hazards of hospitalization such as delirium, urinary incontinence, depression, falls, restraint use, functional decline, adverse drug effects and hospital infections. Therefore, mild pneumonia might be better dealt with inside the long term care facility. In patients with a limited life expectancy (for example, those with advanced dementia), end-of-life pneumonia also requires recognition and appropriate, palliative care.

In immunocompromised patients, prophylaxis with co-trimoxazole (trimethoprim/sulfamethoxazole), atovaquone, or regular pentamidine inhalations may help prevent PCP.

Antipneumocystic medication is used with concomitant steroids in order to avoid inflammation, which causes an exacerbation of symptoms about four days after treatment begins if steroids are not used. By far the most commonly used medication is trimethoprim/sulfamethoxazole, but some patients are unable to tolerate this treatment due to allergies. Other medications that are used, alone or in combination, include pentamidine, trimetrexate, dapsone, atovaquone, primaquine, pafuramidine maleate (under investigation), and clindamycin. Treatment is usually for a period of about 21 days.

Pentamidine is less often used as its major limitation is the high frequency of side effects. These include acute pancreatic inflammation, kidney failure, liver toxicity, decreased white blood cell count, rash, fever, and low blood sugar.

Treatment for "Klebsiella" pneumonia is by antibiotics such as aminoglycosides and cephalosporins, the choice depending upon the person’s health condition, medical history and severity of the disease.

"Klebsiella" possesses beta-lactamase giving it resistance to ampicillin, many strains have acquired an extended-spectrum beta-lactamase with additional resistance to carbenicillin, amoxicillin, and ceftazidime. The bacteria remain susceptible to aminoglycosides and cephalosporins, varying degrees of inhibition of the beta-lactamase with clavulanic acid have been reported. Infections due to multidrug-resistant gram-negative pathogens in the ICU have invoked the re-emergence of colistin. However, colistin-resistant strains of "K. pneumoniae" have been reported in ICUs. In 2009, strains of "K. pneumoniae" with gene called New Delhi metallo-beta-lactamase ( NDM-1) that even gives resistance against intravenous antibiotic carbapenem, were discovered in India and Pakistan."Klebsiella" cases in Taiwan have shown abnormal toxicity, causing liver abscesses in people with diabetes mellitus (DM), treatment consists of third generation cephalosporins.

There is no readily available evidence on the route of administration and duration of antibiotics in patients with pleural empyema. Experts agree that all patients should be hospitalized and treated with antibiotics intravenously. The specific antimicrobial agent should be chosen based on Gram stain and culture, or on local epidemiologic data when these are not available. Anaerobic coverage must be included in all adults, and in children if aspiration is likely. Good pleural fluid and empyema penetration has been reported in adults for penicillins, ceftriaxone, metronidazole, clindamycin, vancomycin, gentamycin and ciprofloxacin. Aminoglycosides should typically be avoided as they have poor penetration into the pleural space. There is no clear consensus on duration of intravenous and oral therapy. Switching to oral antibiotics can be considered upon clinical and objective improvement (adequate drainage and removal of chest tube, declining CRP, temperature normalization). Oral antibiotic treatment should then be continued for another 1–4 weeks, again based on clinical, biochemical and radiological response.

Prevention of VAP involves limiting exposure to resistant bacteria, discontinuing mechanical ventilation as soon as possible, and a variety of strategies to limit infection while intubated. Resistant bacteria are spread in much the same ways as any communicable disease. Proper hand washing, sterile technique for invasive procedures, and isolation of individuals with known resistant organisms are all mandatory for effective infection control. A variety of aggressive weaning protocols to limit the amount of time a person spends intubated have been proposed. One important aspect is limiting the amount of sedation that a ventilated person receives.

Other recommendations for preventing VAP include raising the head of the bed to at least 30 degrees. Antiseptic mouthwashes such as chlorhexidine may also reduce the risk of VAP, although the evidence is mainly restricted to those who have undergone cardiac surgery.

American and Canadian guidelines strongly recommend the use of supraglottic secretion drainage (SSD) Special tracheal tubes with an incorporated suction lumen as the EVAC tracheal tube form Covidien / Mallinckrodt can be used for that reason. New cuff technology based on polyurethane material in combination with subglottic drainage (SealGuard Evac tracheal tube from Covidien/Mallinckrodt)showed significant delay in early and late onset of VAP.

A recent clinical trial indicates that the use of silver-coated endotracheal tubes may also reduce the incidence of VAP. There is tentative evidence that the use of probiotics may reduced the likelihood of getting VAP, however it is unclear if probiotics affect ICU or in-hospital death.

Proven empyema (as defined by the "golden" criteria mentioned earlier) is an indication for prompt chest tube drainage. This has been shown to improve resolution of the infection and shorten hospital admission. Data from a meta-analysis has shown that a pleural fluid pH of <7.2 is the most powerful indicator to predict the need for chest tube drainage in patients with non-purulent, culture negative fluid. Other indications for drainage include poor clinical progress during treatment with antibiotics alone and patients with a loculated pleural collection.

Because of the viscous, lumpy nature of infected pleural fluid, in combination with possible septation and loculation, it has been proposed that intrapleural fibrinolytic or mucolytic therapy might improve drainage and therefore might have a positive effect on the clinical outcome. Intrapleural fibrinolysis with urokinase decreased the need for surgery but there is a trend to increased serious side effects.

Approximately 15 to 40 percent of people require surgical drainage of the infected pleural space because of inadequate drainage due to clogging of the chest tube or loculated empyema. Patients should thus be considered for surgery if they have ongoing signs of sepsis in association with a persistent pleural collection despite drainage and antibiotics. Video-assisted thoracoscopic surgery (VATS) is used as a first-line therapy in many hospitals, although open thoracic drainage remains a frequently used alternative technique.

Treatment of bronchiolitis is usually focused on the symptoms instead of the infection itself since the infection will run its course and complications are typically from the symptoms themselves. Without active treatment half of cases will go away in 13 days and 90% in three weeks.

Measures for which the evidence is unclear include nebulized epinephrine, nasal suctioning, and nebulized hypertonic saline. Treatments which the evidence does not support include salbutamol, steroids, antibiotics, antivirals, chest physiotherapy, and cool mist.

Currently other medications do not yet have evidence to support their use. Ribavirin is an antiviral drug which does not appear to be effective for bronchiolitis. Antibiotics are often given in case of a bacterial infection complicating bronchiolitis, but have no effect on the underlying viral infection. Corticosteroids have no proven benefit in bronchiolitis treatment and are not advised. DNAse has not been found to be effective.

Throughout history treatment relied primarily on β-lactam antibiotics. In the 1960s nearly all strains of "S. pneumoniae" were susceptible to penicillin, but more recently there has been an increasing prevalence of penicillin resistance especially in areas of high antibiotic use. A varying proportion of strains may also be resistant to cephalosporins, macrolides (such as erythromycin), tetracycline, clindamycin and the quinolones. Penicillin-resistant strains are more likely to be resistant to other antibiotics. Most isolates remain susceptible to vancomycin, though its use in a β-lactam-susceptible isolate is less desirable because of tissue distribution of the drug and concerns of development of vancomycin resistance. More advanced beta-lactam antibiotics (cephalosporins) are commonly used in combination with other drugs to treat meningitis and community-acquired pneumonia. In adults recently developed fluoroquinolones such as levofloxacin and moxifloxacin are often used to provide empiric coverage for patients with pneumonia, but in parts of the world where these drugs are used to treat tuberculosis resistance has been described.

Susceptibility testing should be routine with empiric antibiotic treatment guided by resistance patterns in the community in which the organism was acquired. There is currently debate as to how relevant the results of susceptibility testing are to clinical outcome. There is slight clinical evidence that penicillins may act synergistically with macrolides to improve outcomes.

Evidence does not support the general use of antibiotics in acute bronchitis. While some evidence suggests antibiotics speed up resolution of the cough by about 12 hours there is a greater risk of gastrointestinal problems and no change in longer term outcomes. Antibiotics use also leads to the promotion of antibiotic-resistant bacteria, which increase morbidity and mortality.

The Infectious Disease Society of America (IDSA) recommends treating uncomplicated methicillin resistant staph aureus (MRSA) bacteremia with a 14-day course of intravenous vancomycin. Uncomplicated bacteremia is defined as having positive blood cultures for MRSA, but having no evidence of endocarditis, no implanted prostheses, negative blood cultures after 2–4 days of treatment, and signs of clinical improvement after 72 hrs.

The antibiotic treatment of choice for streptococcal and enteroccal infections differs by species. However, it is important to look at the antibiotic resistance pattern for each species from the blood culture to better treat infections caused by resistant organisms.

The treatment of gram negative bacteremia is also highly dependent on the causative organism. Empiric antibiotic therapy should be guided by the most likely source of infection and the patient's past exposure to healthcare facilities. In particular, a recent history of exposure to a healthcare setting may necessitate the need for antibiotics with "pseudomonas aeruginosa" coverage or broader coverage for resistant organisms. Extended generation cephalosporins such as ceftriaxone or beta lactam/beta lactam inhibitor antibiotics such as piperacillin-tazobactam are frequently used for the treatment of gram negative bacteremia.

In cases of viral pneumonia where influenza A or B are thought to be causative agents, patients who are seen within 48 hours of symptom onset may benefit from treatment with oseltamivir or zanamivir. Respiratory syncytial virus (RSV) has no direct acting treatments, but ribavirin in indicated for severe cases. Herpes simplex virus and varicella-zoster virus infections are usually treated with aciclovir, whilst ganciclovir is used to treat cytomegalovirus. There is no known efficacious treatment for pneumonia caused by SARS coronavirus, MERS coronavirus, adenovirus, hantavirus, or parainfluenza. Care is largely supportive.

Corticosteroids, such as dexamethasone and budesonide, have been shown to improve outcomes in children with all severities of croup. Significant relief is obtained as early as six hours after administration. While effective when given by injection, or by inhalation, giving the medication by mouth is preferred. A single dose is usually all that is required, and is generally considered to be quite safe. Dexamethasone at doses of 0.15, 0.3 and 0.6 mg/kg appear to be all equally effective.

Most cases are self-limited and resolve themselves in a few weeks.

Antibiotics do not help the many lower respiratory infections which are caused by parasites or viruses. While acute bronchitis often does not require antibiotic therapy, antibiotics can be given to patients with acute exacerbations of chronic bronchitis. The indications for treatment are increased dyspnoea, and an increase in the volume or purulence of the sputum. The treatment of bacterial pneumonia is selected by considering the age of the patient, the severity of the illness and the presence of underlying disease. Amoxicillin and doxycycline are suitable for many of the lower respiratory tract infections seen in general practice.