Objective To explore the characteristics and influencing factors of family cohesion and adaptability of adolescent bipolar disorder patients， so as to provide reference for clinical treatment of patients with bipolar disorder. Methods The Family Adaptability and Cohesion Evaluation Scale（ FACES） and the Family Assessment Device（ FAD） were used to evaluate the adolescent patients with bipolar disorder and the ordinary adolescent families. The scores of family cohesion dimension were divided into alienation， independence， intimacy and entanglement from low to high， while the scores of adaptability dimension were divided into fossilization， structure， flexibility and confusion from low to high. FAD has seven subscales， including communication， role， problem solving， emotional response， emotional involvement， behavior control and total function. The lower the score of each subscale， the healthier the family is in this dimension. Multivariate logistic regression analysis was carried out to analyze the influencing factors of family cohesion， adaptability and family type by SPSS 20.0 statistical software. Rusults In terms of cohesion， the distribution of alienation， independence， intimacy and entanglement between patient families and healthy families was statistically significant（ Z=-3.42，P ＜ 0.01）. The alienation of patient families was associated with problem solving（ OR=7.24， 95%CI=1.79-29.67） and role score（ OR=3.86， 95%CI=1.63-5.56）. The severity of family alienation increases with the scores of problem solving and role. In terms of adaptability， the distribution of fossilization， structure， flexibility and confusion was statistically significant between patient families and healthy families（ Z=-4.36，P＜0.01）. The fossilization of patient families was associated with age（ OR=1.36，95% CI= 1.09-1.69）， rural area（ OR=0.35， 95%CI=1.96-0.16）， problem solving（ OR=8.08，95%CI=0.55-3.63）， role （OR=58.56，95%CI=0.14-0.85） and behavioral control（ OR=0.15，95%CI=0.03-0.69）. The fossilization of family is more serious when they have higher age and higher scores in problem solving and role dimensions. Compared with rural families， urban families are more rigid and have higher scores of behavior control， and the more likely they are to be confused. In terms of family type， the distribution of extreme， intermediate and balanced type was statistically significant between patient families and ordinary families（ χ2=27.39，P＜ 0.01）. The polarization of patient families was associated with roles（ OR=45.49， 95%CI=2.50-82.99） and emotional intervention（ OR=0.11，95%CI=0.02-0.74）. Conclusions The family cohesion of adolescents with bipolar disorder were affected by role and problem solving， and the adaptability were affected by age， region， problem solving， role and behavior control， which should be considered in psychological intervention.