Bangladeshi women are facing many serious health problems at three stages of childbearing period, although there has been improvement during the recent past, the situation is still worse. In this study, our goal is to find some important risk factors associated with various stages of maternal complications using count data model. We have used Bangladesh Maternal Mortality and Health Care Survey 2010 data in this study and we have extracted a total of 17521 women where our selection criterion was women at age 13-49 and given a live birth from three years prior to the survey. We have applied both univariate and bivariate techniques along with chi-square test statistics in exploratory data analysis and Poisson regression model is used based on the data at three stages of maternal complications namely pregnancy, delivery, postpartum along with potential factors as covariates. We have revealed some important findings from this study. Higher number of parity is positively associated with higher number of complications. Women from Sylhet division are at high risk of facing maternal complications at all stages compared to all other divisions. Education is playing an important role in reducing maternal complication. Poor (poorer, poorest) people are the most vulnerable group among middle and rich (richer, richest) category of wealth index. Previous dismissed pregnancy record has a strong positive association with higher number of complications. We have also showed an important interrelation among treatment seeking behavior related factors (ANC visit number, ANC place, cost at three maternal stages, delivery attendant, delivery place) with higher number of maternal complications.