This appears to be indicitive of excessive inductance in your circuit, either in the circuit wiring or on the scope probe earth lead. You have to be aware that your pulse probably has rise and fall down near a nSec, which means that you are dealing with extremely high bandwidths, so VHF/ UHF design, layout, and measurement techniques are called for. Additionally, the MOSFET needs a high gate drive current to overcome the static and dynamic gate capacitances, if the gate drive is insufficient expect your output waveforms to have excessive rise and fall times. Are the MOSFET and LED capable of handling the rise and fall times desired. Power supply decoupling and very short leads (low inductance) around the LED will be essential. Additionally, constant current drive of the LED may be better. You have not supplied enough information to help much further in my mind, a schematic and the actual circuit layout and measuring technique (an image) would be helpful. Jim Willians, who is no longer with us, had a lot of useful tips on these sorts of issues, his app notes are all available on line, and may be of some helpCheers,Richard