Till a few years ago, malnutrition in children with cancer was persistently ignored. The malnourished state was considered as part of the symptom complex of progressive and active cancer. Only recent it is recognized that malnutrition is closely related to late stage of malignancy and is a poor prognostic sign. In fact it is a frequent cause of death in children with cancer. Compared with adults, children are at a greater risk for nutritional depletion since they have a more rapid metabolic rate and greater caloric needs for growth and development. In this thesis, my aim is to pursue an active investigation of effect malnutrition on children with malignancy using meta analysis. The studies I have included here were conducted in the Department of Pediatric Haemato-Oncology, Bangabandhu Sheikh Mujib Medical University (BSMMU). Here initially children were categorized as malnourished using different nutritional parameters in different studies. Our aim is to combine these experiments. Meta analysis is a statistical method to combine evidences or results obtained from individual trials. Here some well known fixed effect methods of meta analysis have been discussed. Inverse variance method is applied in this study to combine effect size such as Odds ratio and Relative risk. Several random effect methods have been discussed in this paper along the fixed effect method. In this paper six studies have been included where we find out the status of malnourished children with malignancy with respect to the well nourished children group with malignancy after giving treatment. Meta analysis has been performed to find out the combined effect of malnutrition using these six experiments.