To evaluate the operational effectiveness of equipment using less number of tests, an analyzing and predicting method is proposed by combining orthogonal design, variance analysis and multivariate regression. Orthogonal design is used to select the combination of battlefield condition factors associated with operational effectiveness. The significance formula of factor is derived using variance analysis. An operational effectiveness prediction model with confidence interval is proposed based on multivariate regression. Taking the completion rate of a certain UAV reconnaissance task as an example to illustrate the calculation and analysis process. The results show that, the original 32 times of tests are compressed to 8 tests. The main battlefield condition factors can be accurately identified. The predicted error of task completion rate was only 0.0259. And the accuracy of the prediction interval is 91.67%. The requirement is satisfied for analysis and evaluation of operational effectiveness under small samples.