A statistical study, including discriminant analysis, was carried out on a suite of 438 formation waters from Alberta, Canada, which had been analyzed for Cl, Br, I, HCO₃, SO₄, Ca, Mg, and Na. The analyses were divided into two populations, depending on whether the initial status of the stratigraphic interval in that well was producing oil and/or gas (316 samples) or was nonproducing, i.e., abandoned (122 samples). The populations were subdivided further into Paleozoic and Mesozoic groups, and these two groups, together with the entire suite of analyses, were subjected to statistical study. With 95-percent assurance, the Alberta formation waters associated with producible hydrocarbons are chemically different, in a multivariate sense, from the format on waters from nonproducing stratigraphic intervals. Iodide and magnesium are the most important discriminators in the Paleozoic group. Formation waters from Paleozoic strata have a general background of iodide (and bromide) from argillaceous rocks (including associated organic matter). The presence of source sediments and the resultant hydrocarbon accumulations are accompanied by enhanced contents of iodide in the associated formation waters as a result of the large contents of organic matter in the source sediments. The following indirect but inferred relation may explain the importance of magnesium to the discrimination: Mg enrichment ^rarr increased dolomitization ^rarr increased effective porosity ^rarr increased probability of hydrocarbon presence. In the Mesozoic group the most im ortant discriminators are chloride and sodium. These reflect the fact that in the Mesozoic sequence in Alberta, which consists of only 60 percent marine strata and is mostly within the regime of local and intermediate flow systems, the most probabilistic situation for both hydrocarbons and high-salinity formation water is within the marine-clastic section in regions subject to the minimum hydrodynamic flushing. In such situations the reservoired crude oils have been preserved from bacterial degradation and water washing by fresh waters recharged from the surface. The computed discriminant score may be utilized to indicate the probability of a formation-water sample being associated with hydrocarbons in the subsurface. The acquisition of a larger data base and statistical analysis at the ormation level will increase the probability of successful classification.