Rural Credit Bank (BPR) is a financial institution that has an intermediation function with the activity of collecting funds from the public, in the form of Savings and Time Deposits, and channeling them back to the public in the form of credit. Businessly, the purpose of BPR is to make a profit. BPR's efforts to obtain profit are faced with problems, namely external problems in the form of unfavorable economic conditions and internal problems in the form of credit risk as indicated by the high Non-Performing Loans. This study aims to analyze the macroeconomic influence that is represented by BI Rate, Inflation, GRDP and Export Growth as well as Non-Performing Loans on BPR ROA in Lampung Province. This study uses multiple linear regression analysis tools with the Ordinary Least Square (OLS) method carried out using a time span from 2007 to 2017 and hypothesis testing uses t-statistics to test the partial regression coefficients and the significance of the overall effect with a level of significance of 5%. Because the data used are secondary data in the form of time series data that has wide fluctuations or instability, then testing ARCH (Autoregressive Conditional Heteroscedasticity) and GARCH (Generalized Autoregressive Conditional Heteroscedasticity) and to determine the accuracy of the model need to be tested on several classic assumptions that are underlying regression model. Testing the classic assumptions used in this study include tests of normality, heteroscedasticity, autocorrelation, and multicollinearity. During the observation period showed that the research data were normally distributed. Based on the heteroscedasticity test and the multicollinearity test, no variables found that deviate from the classical assumptions, but based on the autocorrelation test found a positive autocorrelation. This autocorrelation problem is most likely due to the small amount of data (n). Overall this shows that the available data meets the requirements to use the multiple linear regression equation model. Based on the test results the coefficient of determination obtained R2 value of 0.5164 which means that the closeness of the overall independent variable to the dependent variable is 51.64%, while the remaining 48.36% is influenced by other variables outside this regression model. Based on the F statistical test at the 95% confidence level, the calculated F value is 1.28 and the F-Prob value is 0.3805> α 5%, so it can be concluded that the overall variables of BI Rate, Inflation, GRDP, Export Growth and NPL influence ROA. Based on the t test it was concluded that the BI rate, inflation, GRDP and NPL did not have a significant negative effect on ROA, while export growth had no significant positive effect on ROA.