Petroleum hydrocarbons inevitably are released into the environment and contaminate surface water bodies particularly near production or storage sites, but also escape accidentally during handling, transport or processing. Accurate measurement of dispersed oil in water is thus an important task which is usually done by gas chromatographic methods. The two standards that are typically followed in this respect are the DIN ISO 9377-2:2000 standard and the OSPAR (Oslo-Paris commission) method. In contrast to the DIN ISO method, the analytical method proposed by the OSPAR Analytical Method Committee is based on large volume injection (LVI) of the petroleum hydrocarbon extract in order to avoid any external preconcentration step, but still reach the stipulated detection limit. Although a standard method should provide a sufficient degree of robustness, we demonstrate here that the correctness and precision of results for the hydrocarbon index strongly depends on the proper choice of measurement parameters whose correct selection is left to the judgement of the analyst. Moreover, even under standardized conditions of measurement, the results will always show a sample-specific bias. In this work we discuss the influence of these parameters on the results, and what challenges the determination of petroleum hydrocarbons in water poses in general. The inadequacy of the two methods to provide a full equivalent to the earlier ‘total petroleum hydrocarbon’ (TPH) method is highlighted using a number of typical oil samples of different origin as example.