In order to reduce engine out CO2 emissions, it is a main subject to find new alternative fuels from renewable sources. For identifying the specification of an optimized fuel for engine combustion, it is essential to understand the details of combustion and pollutant formation. For getting a better understanding of the flame behavior, dynamic structure large eddy simulations are a method of choice. In the investigation presented in this paper, an n-heptane spray flame is simulated under engine relevant conditions starting at a pressure of 50 bar and a temperature of 800 K. Measurements are conducted at a high-pressure vessel with the same conditions. Liquid penetration length is measured with Mie-Scatterlight, gaseous penetration length with shadowgraphy and lift-off length as well as ignition delay with OH*-Radiation. In addition to these global high-speed measurement techniques, detailed spectroscopic laser measurements are conducted at the n-heptane flame. Spontaneous raman scattering is used to measure temperature and oxygen concentration, while laser-induced fluorescence is applied to measure quantitative nitrogen concentrations. The LES-simulation matches the parameters liquid penetration length, gaseous penetration length and lift-off-length as well as ignition delay very well. Concerning the detailed flame parameters it can be seen in the measurements that adiabatic flame temperatures are achieved in the middle of the flame kernel. The NO concentrations measured in the jet core can be compared to 0-d-constant-pressure simulations, taking the residence time of the fluid parcels in the main reaction zone into account. When these measurements are compared to the results of the LES-simulations, it is obvious that the turbulence formation in the simulation is different from the measurement. A higher amount of oxygen is predicted in the flame kernel and the adiabatic flame temperature is not reached in the middle of the flame.