This article investigates the patterns and determinants of family planning (FP) communication between husbands and wives in Bangladesh using a couple dataset from the 1999-2000 Demographic and Health Survey. Nearly 50% of the couples reported not discussing FP matters between each other and less than 10% reported discussing FP more than twice during year preceding the survey. Results from the two-level random intercept binary logistic regression analyses suggest that the odds of FP communication are considerably high among couples in rural areas, those who have five and more living children, where both the spouses are educated, both approve of FP and currently using a method, when the husband has access to newspapers and where husbands aged less than 40 years and wives aged less than 35 years and couples had spent less than 11 years in marriage. Community variations within the PSU (Primary Sampling Unit) level were found significant which suggest that inter-spousal communication has not been established as an expected cultural behaviour uniformly across different communities.