Breast cancer is the most frequently occurring cancer next to cervical cancer in both developed and developing countries with high risk in developed countries and low risk in developing countries. The main objective of the study is to investigate the risk factors, median survival time, compare the survival curves, compare the performance of AFT and parametric frailty models in modeling survival time of breast cancer patients. A retrospective study was employed on 392 breast cancer patients registered from 2013-2018 in University of Gondar and Felege Hiwot Referral Hospitals in northwest of Ethiopia. The median survival time of breast cancer patients who live in the urban and rural areas are 34 and 35 months, respectively. The family history of breast cancer has 0.643 times shortened survival time than no family history of breast cancer. The frailty of University of Gondar and Felege Hiwot Referral Hospitals were 0.536 (µ < 1) and 1.465 (µ > 1), respectively. The study employed multivariable lognormal-gamma shared frailty model and the results of the model revealed that age, weight, place of residence, tumor size, stage, number of estrogens, status of estrogen receptors, number of progesterone and family history of breast cancer are statistically significant factors which affect the survival time of breast cancer patients.