In the development of a new diagnostic motor performance test to spare more children from painful muscle biopsy, seven functional items were used to measure muscle strength and muscle endurance in a prospective study on new patients. Over a 2-year period, 22 patients (12 males, 10 females; mean age 8y 1mo [SD 2y 6mo], range 4–11y) were recruited for the study. They had all been referred with suspected myopathy. The motor performance test was administered before muscle biopsy. Validity of the seven items was assessed using logistic regression analysis and receiver operating characteristic (ROC) analysis. Two items were withdrawn from the test because they were not suitable for children aged 4 to 5 years. The five remaining items were: Heels, Circuit, Stairs, Jump, and Gowers. A full logistic regression model including these five items was fitted to the total population of 90 patients suspected of having myopathy (from this study and our previous study) to make the best prediction of whether myopathy was present. The ROC area under the curve of the diagnostic prediction model was 0.92 (95% confidence interval [CI] 0.87–0.98) and 0.89 (95% CI 0.87–0.92) after bootstrap correction. This indicated the high diagnostic power that can be expected for future, similar patients. This non-invasive and child-friendly motor performance test can improve diagnostic procedure and, therefore, spare more children from unnecessary muscle biopsy.