PurposeThe paper examines the effect of bank-specific, financial structure and macroeconomic factors on the profitability of banks in Asian economies during 1995–2017.Design/methodology/approachIt uses the data of 2,446 banks across 47 Asian countries between 1995 and 2017 (41,582 year observations). The static and dynamic panel generalized methods of moments (GMM) estimation techniques are applied.FindingsThe results show that banks that are highly dependent on nontraditional activities have lower net interest revenue and net interest margin but higher return on assets, return on equity and profit before tax. Higher opportunity cost, capitalization, demand deposits and market risk result in a better bank profits. Furthermore, banks with higher loan exposure and growth have more profit. However, nonperforming loans have negative and significant impact on bank profitability. Asian banks do not suffer from diseconomies of scale and scope. The author also finds that banks located in countries with high gross domestic product, inflation rates and high rates of interest or in financially developed economies offer better profits. High credit to the private sector reduces the bank profitability. This study finds evidence to support the structure-conduct-performance (SCP) hypothesis. It also provides evidence that the impact of financial turmoil on the profitability of the Asian banking sector is negative and significant and has severely weakened the Asian banking system.Originality/valueAs Asia has become an important economic area and the Asian topic has not earned enough discussions, this paper is the first to examine Asian banks with the latest and a wider range of panel data that cover 2,446 banks at 47 Asian countries over the period 1995–2017. The present study is among the first to address the influence of financial turmoil on bank profitability in this region. It also studies new variables, such as demand deposits, opportunity cost and off-balance sheet activities, which have not been examined in relation to bank profitability. It also applies both static techniques and dynamic panel estimation techniques to analyze the data.