OBJECTIVE： To investigate the causes of the variation of clinical inpatients in inpatients with community acquired pneumonia （CAP） of Dongfang Hospital （called “our hospital” for short）， and analyze drug use of clinical pathway. METHODS： The medical records of 255 CAP variant patients in general wards （non-ICU） were collected from HIS database of our hospital during Jul. 2015-Jun. 2016. General information， variation cause， medication plan and drug use were analyzed statistically by using Excel 2016 software. RESULTS： Among 255 patients， there were 142 female cases （55.69％） and 113 male cases （44.31％）. Most of cases occurred under the age of 65 （88.24％）. Hospitalization duration mostly ranged 7-14 d （69.02％）. Totally 98 patients were less than 65 years old and had no complications （37.65％）. Variation causes were divided into uncontrollable causes （99.32％） and controllable causes （0.68％）. The uncontrollable causes included disease outcome （70.75％）， patient demand （26.87％） and other causes （1.70％）； disease outcome mainly included increasing medical order due to the change of disease condition （37.41％）； patient demand mainly included patients or their family members refused to （or required） discharge （13.95％）. The controllable causes contained hospital system （0.34％） and medical staff （0.34％）. There were 11 kinds of drugs in the clinical pathway medication plan templates of CAP in our hospital， among which one kind of anti-inflammatory drug lacked of evidence in Guidelines for Diagnosis and Treatment of Adult Community Acquired Pneumonia in China （2016 edition）； 2 kinds of drugs （penicillin， clindamycin） were not mentioned in modified edition of American Society of Infectious Diseases/American Thoracic Society Diagnosis and Treatment Guidelines for Adult Community Acquired Pneumonia （2009）. The CAP variant patients aged below 65 years old without complications were mainly given quinolones， macrolides or β-lactams alone， accounting for 56.25％. The CAP variant patients aged below 65 years old with complications and aged 65 years old or above were given quinolones or carbapenems alone， accounting for 47.34％. Quinolones+β-lactams/cephalosporins accounted for 41.42％. CONCLUSIONS： The causes for the variation of CAP clinical pathway are mainly uncontrollable causes such as disease outcome. There is a lack of clinical evidence support in clinical pathway medication plan and implementation. It is suggested  to strengthen administrative management according to variation causes，analyze and deal with them in time，and modify and perfect the medication template plan of clinical pathway. It is suggested to formulate medication plan according to age and disease condition， so as to further standardize CAP medication plan and guarantee the effects of clinical pathway.