Indirect effects of lightning refer to upset of or damage to electrical equipments and systems. Evaluation of aircraft susceptibility to indirect effects of lightning is usually performed through experimental tests. A current homothetic to the real threat but with a lower amplitude is injected through the aircraft structure. Analysis of aircraft transfer functions is carried out by measuring induced current and voltage and extrapolating the results to the standardized threat. The extrapolation process may influence the results especially if non linear effects appear during the test.An analysis of non linear effects has been carried out in the framework of the European FULMEN program and is part of the global definition of aircraft interaction with lightning at the equipment level (internal threat).The different sources of non linearity are identified : arcing and sparking, non linear protection devices, non linearity of material and influence of the extrapolation method. An experiment has been designed and performed at CEAT in May 1998 on three structures: an unprotected composite structure, a composite structure protected with a copper mesh and a metallic cylinder. An increasing current level is injected into each structure and sparks appear naturally or are artificially created. The analysis is made by measuring voltage and current on wires inside the structure and making comparison between high level injections and extrapolated levels.