A technique is applied to ohtain unbiased eslimators with minimttm variance among the class of linear estimators for the standard deviation of of logistic. population with known mean. Coefficients for multiplying ordered observations are developed for complete and completc samples forn= 2(1)26. Each sample of sizenis censored from above wherem≤n. Then the smallest subset oflorder slatistics from the set ofmavailable order statistics is found which yields at list a 98.5% efficiency relative to them-order-statistic estimator. The coefficients for thel-order-statistic estimator, the variance and efficiency relative to the estimatot, using allmorder statistics are persented.