Evidence for on site biodegradation may be difficult to provide at heterogeneous sites without additional experiments in controlled laboratory conditions. In this study, microbial activities measured as CO2 and CH4 production were compared in situ, in intact soil cores and in bottle microcosms containing sieved soils. In addition, biodegradation rates were determined by measuring the decrease in petroleum hydrocarbon concentrations at 7 degrees C in aerobic and anaerobic conditions. Elevated concentrations of CO2 and CH4 in the soil gas phase indicated that both the aerobic and anaerobic microbial activity potentials were high at the contaminated site. Aerobic and anaerobic microbial degradation rates in laboratory experiments of petroleum hydrocarbons were highest in soils from the most contaminated point and degradation in the aerobic and anaerobic microcosms was linear throughout the incubation, indicating mass-transfer-dependent degradation. Different results for microbial activity measurements were obtained in laboratory studies depending on pretreatment and size of the sample, even when the environmental conditions were mimicked. These differences may be related to differences in the gas exchange rates as well as in changes in the bioavailability of the contaminant in different analyses. When predicting by modeling the behavior of an aged contaminant it is relevant to adapt the models in use to correspond to conditions relevant at the contaminated sites. The variables used in the models should be based on data from the site and on experiments performed using the original aged contaminant without any additions.