Steady state mapping is fundamental to optimizing IC engine operation. Engine variables are set, a predefined settling time elapses, and then engine data are logged. This is an accurate but time consuming approach to engine testing. In contrast the sweep method seeks to speed up data capture by continuously moving the engine through its operating envelope without dwelling. This is facilitated by the enhanced capability of modern test rig control systems. The purpose of this work is to compare the accuracy and repeatability of the sweep approach under experimental conditions, with that of steady state testing. Limiting factors for the accuracy of the sweep approach fall into two categories. Firstly on the instrumentation side - transducers have a characteristic settling time. Secondly on the engine side - thermal and mechanical inertias will mean that instantaneous measurements of engine parameters differ from the steady state values. These errors can be reduced to satisfactory levels by modelling the engine and instrumentation responses. Some data i.e. air, fuel and emissions pose significant problems, with the response time of the equipment limiting the maximum sweep speed. Despite requiring intensive data processing and test bed sophistication, sweep mapping shows the potential to give accuracy comparable to steady state testing, but with much reduced mapping times.