Engine control optimization, with its always growing complexity, is in permanent focus of engine researchers and developers all over the world. Automotive engines are dominantly used in dynamic conditions, but generally, steadystate operating points are used for building up mathematical models which are later subject to the numerical optimization. For this purpose, a large amount of steadystate regimes needs to be evaluated through experimental work at the engine test stand, which is an extremely time and funds consuming process. Consequently, the methodology for data gathering during engine dynamic excitation could lead to significant savings at the expense of acceptable data accuracy loss. The slow dynamic slope method starting from a stationary operating point was evaluated by several authors in the past. In this paper, slow dynamic slope method with exclusively transient excitation will be presented drawing attention to some of its advantages and drawbacks. The rate of change of engine load as a main control parameter during dynamic test is of great importance for the quality of the final data and for total test duration. In this regard, several tests of different duration were applied for fixed engine speed values to cover engine speed-load usage domain. An approximation of stationary testing results obtained in this way could be used for evaluation of the map gradients and thus as a guide-line for additional stationary tests based on design of experiment method.