The selection of an activity test to screen plants for antimicrobial agents is in itself a problem. However, the preparation of those extracts is a very important step and full attention has not yet been given to this issue. This study was performed in two parts. In the first, different solvents were tested to select the best extractant for three plants used in Mexico to treat infections: Leucophyllum frutescens, Cordia boissieri, and Cyperus alternifolius. In the second part, an experimental design was developed for the most active extract assayed in the first part of the study, using factorial analysis at two levels. The variable under evaluation was the minimum inhibitory concentration (MIC) of the extract active against a resistant strain of Staphylococcus aureus, and the number of active spots identified in a bioautographic assay. The solvents used for the first phase of the study were selected from those most often cited in articles on antimicrobial screening. The best recoveries and lowest MICs were obtained with hydroalcoholic extracts. Shorter extraction times and storage times gave lower MICs; however, the volume of solvent and speed of agitation had no significant effect on MIC.
