Lipids are markers of living as well as non-living organic matter in soils. Lipids may be of microbial, faunal, plant and even human origin. They constitute, in many ways, the long, medium and short term memories of what happens in the soil. Discovering how to unravel these memories help us understand the changes in the soil due to environmental (rain, temperature, plant type, etc.) and anthropogenic impacts (tillage, pollution, etc.). A better understanding of the roles of lipids in soil could provide information about carbon sequestration potential that may help us understand the impacts of various soil management strategies with regard to improving crop productivity and develop strategies for bioremediation of polluted soils. Separating and measuring lipid components with classical techniques is both difficult and complex. A rapid analysis technique, pyrolysis Molecular Beam Mass Spectrometry (py-MBMS), was developed for analyzing the organic matter content and chemical composition of whole soils and soil extracts. Py-MBMS has several advantages, such as speed of analysis, sensitivity, high sample throughput, minimal sample preparation, and low cost. We have used pyrolysis molecular beam mass spectrometry (py-MBMS) to specifically fingerprint the lipid composition (signature) of five agricultural soils. Since the lipids in soils originated from living and dead biota and stabilized organic matter, py-MBMS is used to obtain a ‘classification' of soils according to their biotic content and to detect specific lipid biomarkers. The soils used in this study are all from the Crete soil series. Four cores (at 0-15 cm depth) were sampled in different cultures (soybean, corn, alfalfa, replanted prairie and native grass) in summer 2005. Soil samples were rapidly heated (100-500 mg) in an inert, helium atmosphere at 550°C. The 0.5 g sample size improved instrument sensitivity for samples containing <0.2-wt% carbon. The pyrolysis products were sampled directly in real time and introduced into the ionization region of the mass spectrometer. Mass spectrometry provides universal detection of all sampled products and the molecular beam sampling ensures that representative products are detected. The py-MBMS method is rapid (1-5 minutes) and can analyze up to 150 samples per day. Multivariate statistical analysis of the py-MBMS data can determine the proportion of the Soluble Organic Carbon (SOC) found in different pools, such as, carbohydrate, lignin, fatty acids and humic acids. Recent work shows that the method can distinguish both sample depth and management practice of agricultural soils. In this work, we applied these techniques to the soils described above to determine if we could see lipid species. This sample set included triplicate analyses per sample with a soil reference material (3.0 wt% SOC) used for instrument monitoring every tenth sample. The resultant pyrolysis mass spectra of the native samples (10-500 amu) are chemically rich and very complex. We use multivariate data analysis (pattern recognition) to handle this large data set and identify trends to discover the underlying chemical changes that may not be obvious by comparison of such complex mass spectra. Principal component analysis of the mass spectral data from the whole soils shows that the native grass and replanted prairie soils are distinguishable from cropped soils. 90% of the variance is explained by principal component 1 (soil carbon content). Native and replanted soils are more similar chemically and have more carbon than the cropped soils. Crop type seems to result in slight chemical differences. Ergosterol, a fungal lipid biomarker, appears in all most of all samples but the signal is low. Py-MBMS analysis of the dried lipid extracts are complex and chemically rich. The extract mass spectra are chemically distinct for the varied soils used in this study. We are currently working on identifying the lipid components in these samples. Results on the lipid signatures in the soil lipid extracts will also be presented in this communication. Our long term goal is to develop effective mass spectrometry tools for detecting and quantifying (fingerprinting and profiling) simple and complex lipid molecules in soils.