The aim of this study was to assess an optimal screening for paediatric patients suspected of mitochondriocytopathy to justify a muscle biopsy. Forty-five patients were included. Medical history, physical examination, cardiac and ophthalmologic evaluation, clinical chemical investigations, in vivo function tests, neuroimaging and a skeletal muscle biopsy were performed in all patients. The results of the biochemical muscle studies were compared with the results of the other investigations. First, parameters with a statistical relationship with the result in muscle, normal or deficient, were selected. Secondly, a prognostic index was constructed using these parameters. Five parameters were selected: age <4 years, elevated fasting lactate to pyruvate ratio, elevated thrombocyte count, elevated lactate, and elevated alanine. Each parameter was scored 0 (not present) or 1 (present). The chance of a normal biopsy with a given value of this index (sum of the scores) was calculated: logit (Pr) = alpha + beta x index; alpha: -0.8167 and beta: 0.8331. (Pr: probability of normal biopsy.) The chance of a normal biopsy with an index value of 5 is 0.03, 4 is 0.07, 3 is 0.16, 2 is 0.30, 1 is 0.50 and 0 is 0.69. This prognostic index is a valuable instrument in deciding whether the suspicion of mitochondriocytopathy is strong enough to merit a muscle biopsy.