Sore throat is the second most common cause of antibiotic prescribing in primary care in Sweden. Guidelines for sore throat focus on identifying people with sore throat where there are 3 and 4 specified criteria and where near patient tests identify group A streptococci (GAS). In these cases, phenoxymethylpenicillin is recommended. Studies that have identified microorganisms in sore throat show that there are other bacteria and viruses than GAS, that give similar symptoms and that sometimes no microorganism is trapped despite pronounced symptoms. In recent years, a bacterium F. necrophorum has been identified, which is found in increased incidence of sore throat, but it is also found in healthy individuals. In clinical practice, many patients are treated with penicillin even if GAS is not captured. This may be because the doctor perceives the patient as sick or because other bacteria are not caught with a near patient test which causes the doctor to treat anyway. The specific aims are to in patients with GAS-negative sore throat and 3 and 4 criteria, aged 15 years and older in primary care, study whether phenoxymethylpenicillin treatment shortens the duration of the disease, reduces the symptom intensity and sickness absence, and investigates the importance of other microorganisms than GAS in sore throat. The study is a randomized controlled trial in which patients with sore throat are randomized to phenoxymethylpenicillin 3 times daily for 10 days or to no antibiotic therapy. There will also be and a reference group with severe (Centor score 3-4), GAS-positive acute tonsillitis. Blood samples for inflammatory and immunological response to infections are taken. Throat samples for culture of F. necrophorum and streptococcal groups C and G, as well as polymerase chain reaction (PCR) analysis for bacteria and viruses are also taken at inclusion and at follow-up. The outcome will be followed in a patient diary for 10 days and at a return visit after 18-24 days where the clinical outcome is asked for and where the blood- and throat samples are repeated. Follow-up will also takes place via e-mail after 1 and 3 months.