Quantitative structure-activity relationships (QSARs) for estimating the adsorption of organic chemicals on activated carbon were developed using molecular connectivity. The QSARs were based on adsorption data for 363 chemicals [expressed as (X/C)(min)] obtained from six literature sources, representing a wide range of aromatic and aliphatic chemical classes. In order to test the predictive capabilities of the QSAR equations and to illustrate their use, the QSARs were applied to data from three additional data sources. The existence of divergent literature values for the same chemicals was addressed in one QSAR by a ''dummy variable'' to identify the data source. The resulting QSARs are well suited for use in the water treatment industry because they are based on a wide range of chemicals, use readily calculated parameters, and represent adsorption at low adsorbate concentrations. However, in applying QSARs based on literature data, it is imperative to consider the relevance of the data source to one's own investigative situation.