Improper disposal of wastes by former manufactured gas plant (FMGP) sites has resulted in serious environmental threats today. Some of the main contaminants from MGP activities include tar, benzene, toluene, ethylbenzene and xylenes (BTEX), polycyclic aromatic hydrocarbons (PAH), metals and cyanides. During the last decade, monitored natural attenuation (MNA) has become one of the widely accepted choices among researchers and regulators for the clean up of such contaminated sites. The objectives of this study were to develop protocols for the use of Geoprobe direct push technologies (DPT) to characterize the geology and hydrogeology at FMGP sites, compare the capabilities of several publicly available groundwater flow and contaminant transport models (BIOSCREEN, BIOPLUME III and MODFLOW/MODPATH/RT3D) and select a model with the best overall capabilities to model groundwater flow, contaminant transport and to estimate the field-scale biodegradation rates for the implementation of natural attenuation at the Cherokee, Iowa FMGP site. DPT pre-packed screen monitoring wells and dual tube direct push equipment were found to provide results that were equivalent to conventional 2-inch monitoring wells for groundwater sampling and hydraulic conductivity. Use of DPT electrical conductivity probes allowed geological stratigraphy to be easily mapped and provided detailed information of a pinch zone that restricted the groundwater flow at the site. Comparison of different groundwater and transport model showed MODFLOW/MODPATH/RT3D is a more versatile model due to its ability to model the flow and transport in 3-D and provide stable groundwater head and contaminant concentration results. Biodegradation rates were computed using analytical and numerical (modeling) methods for selected compounds. Biodegradation rates for BTEX and PAH estimated by analytical method varied from 0.00019 d−1 and 0.0022 d−1 and 0.0003 d−1 to 0.0003 d−1, respectively. Biodegradation rates calculated for toluene and phenanthrene by numerical method were 0.03 d−1 and 0.006 d−1, respectively. Rates calculated by these two methods were different because of incomplete source characterization and insufficient groundwater sampling locations. A sensitivity analysis of the RT3D model showed that biodegradation rate is the most sensitive parameter.