This paper reports significant new findings on banking performance by relating (i) 7 bank-specific measures, (ii) 3 non-bank-specific measures with (iii) time dummy variable as control for financial crisis in the test period. The findings concern the performance of a new type of bank called the 'participation bank'. Participation banks price their funding through profit-sharing contracts with customers, so deposit and lending costs are decided not on the basis of market interest rates as in the case of mainstream banks. A sample of 100 participation banks covering 25 countries were selected for this study over the financial years 2007-2015. We used a new measure equivalent to the net interest margin called 'profit share margin', which has not been previously used to study banks. In fact, no study using participation banks has been carried out as yet. The dynamic panel GMM procedure was applied to obtain robust estimators; this is a refined econometric method that is also seldom applied in banking studies. The results revealed that 6 bank-specific factors statistically significantly affected the performance of participation banks in the test period. The paper also reports that the practice of including non-bank-specific factors as possibly relevant for performance is questionable as these were not found to be significant. The findings were from both OLS and GMM panel regressions providing comparison statistics with some past studies.