Ingestion of extracts prepared from various medicinal plants to induce or augment labour is common amongst Black South African women during the late stages of pregnancy. This applies particularly to the rural areas where modern health care facilities are often lacking. Many of these plants have not been investigated scientifically and one needs to substantiate claims of quality, safety and efficacy. Furthermore, it is believed that the consumption of these plant extracts can result in foetal meconium staining at delivery. An investigation into the uterotonic properties of three plants viz. Ekebergia capensis Sparrm. Clivia miniata (Lindl.) Regel. and Grewia occidentalis L. were carried out using guinea pig uterine smooth muscle in vitro. Supercritical fluid extraction was performed with water modified supercritical carbon dioxide to extract the uterotonic components. An attempt was also made to couple supercritical fluid extraction directly on-line to the bioassay so that on line screening of crude plant extracts could be performed within short periods of time. The effects of supercritical CO2 decompression on temperature and pH of the muscle bathing solution were considered since these factors affect muscle contractility. The direct effects of excess CO2 on intracellular mechanisms were eliminated by constructing a CO2 reduction interface together with passage of carbogen which aided in the rapid displacement of excess CO2, As samples of these extracts were found to induce muscle contraction, supercritical fluid fractionation (SFF) was performed by sequentially increasing the fluid density. Extracted fractions were obtained by sequentially increasing the pressure at constant temperature and modifier concentration in an attempt to identify the active fractions. Extractions were performed at 200 atm, 300 atm and 400 atm respectively. Subsequent testing of these fractions enabled the detection of active and inactive fractions as well as a fraction that had a spasmolytic effect on uterine muscle. The 400 atm extracts of E. capensis and C. miniata displayed maximum activity while only the 300 atm extract of G. occidentalis induced uterine muscle contraction. Subsequent analysis of the sequentially extracted fractions, by high performance liquid chromatography and micellar electrokinetic capillary chromatography revealed that certain compounds present in the fractions that stimulated muscle contraction, were sensitive to the extraction pressure hence making it possible to determine the compounds that were likely to be active. Column chromatography followed by various spectroscopic techniques were performed in an attempt to isolate and elucidate the structures of the compounds that were present in the plant extracts. The extract of Ekebergia capensis yielded five known compounds (B-sitosterol, oleanonic acid, 3-epioleanolic acid, 2,3,22,23-tetrahydroxy-2,6,1 0, 15,19 ,23-hexamethyl-6, 10, 14, 18- tetracosatetrene and 7-hydroxy-6-methoxy coumarin. The extract of Clivia miniata yieded linoleic acid and 5-hydroxymethyl-2-furancarboxaldehyde while the extract of Grewia occidentalis yielded 3-(4-hydroxy-3-methoxyphenyl)-2-propenal, a novel compound 2,2' ,6,6'-tetramethoxy-4'-al-4-(w-oxo-E-propenyl)-biphenyl and oleanonic acid. The pure compounds were further evaluated pharmacologically to identify the active components and assess the physiological mode of action by the use of various receptor blockers. Oleanonic acid, 3-epioleanolic acid, linoleic acid and 5- hydroxymethyl-2-furancarboxaldehyde and 3-(4-hydroxy-3-methoxyphenyl)-2-propenal were found to induce an agonistic muscle response. All these compounds were observed to mediate their effects through the cholinergic receptors. The results obtained in this study supports the claim of these plants possessing uterotonic properties. / Thesis (Ph.D.)-University of Natal, Durban, 1997.