This paper proposes two measures on economic slackness, the survey-based output gap and the survey-based input gap. Our method extends Carlson and Parkin’s (1975) method of transforming qualitative data into quantitative data and uses Hirose and Kamada’s (2003) method of estimating a time-varying model. We show theoretically how a survey asking firms whether business conditions are favorable or unfavorable generates an estimate of the output gap in a country. Our method is also applicable to the estimation of the capital input gap and the labor input gap, which are combined to produce the survey-based input gap in a country. The Japanese survey-based output gap and input gap are estimated, based on the TANKAN survey conducted by the Bank of Japan. The estimated gap measures are consistent with the official reference dates of business cycle and have tight confidence intervals. The two measures surprisingly coincide with each other, but sometimes deviate from each other, particularly in severe recessions. The deviation is likely to be caused by changes in total factor productivity.