The study examines the degree of concentration and performance of the Bangladesh banking industry for the period 1999-2011 by using the random effects (RE) estimator. It applies two competing hypotheses of the traditional industrial organisation theory e.g. the structure conduct performance (SCP) paradigm and efficient structure hypothesis (ESH) to investigate the relationship between the concentration and competition in the banking sector. The results of the main sample (1999-2011) do not find any support for either of the hypotheses. However, a sub-sample (2002-2011) of the study supports the SCP hypothesis that the profitability of Bangladesh banking market is determined by concentration and not by the market share of banks. It implies that concentration lowers the cost of collusion between banks and results in higher than normal profits for all market participants. Bank performance is positively associated with capitalisation, liquidity and assets size of the banks. The ownership variable suggests that government-owned banks are less profitable than other commercial banks in the market.