1. A prolyl-s-RNA synthetase (prolyl-transfer RNA synthetase) has been purified about 250-fold from seed of PhaseoluB aureus (mung bean), a species not producing azetidine-2-carboxylic acid, and more than 10-fold from rhizome apices of Polygonatum multiflorum, a liliaceous species containing azetidine-2-carboxylic acid. The latter enzyme was unstable during ammonium sulphate fractionation. 2. The enzymes exhibited different substrate specificities towards the analogue. That from Phaseolus, when assayed by the ATP-PP1 exchange, showed azetidine-2-carboxylic acid activation at about one-third the rate with proline. Both labelled imino acids gave rise to a labelled aminoacyl-s-RNA. The enzyme fromPolygonatum, however, activated only proline. 3. The enzyme from Polygonatum also formed a labelled prolyl-s-RNA with Pha8eolu8 s-RNA but at a lower rate than when the Phaseolu8 enzyme was used. No reaction occurred when the Phaseolus enzyme was coupled with Polygonatum s-RNA, and only a very slight one was observed when both enzyme and s-RNA came from Polygonatum. 4. Protein preparations from seeds of Pisum sativum, another species not producing azetidine-2-carboxylic acid, also activated the analogue in addition to proline, whereas those from rhizome and seeds of Convallaria, the species from which the analogue was originally isolated, failed to activate it. However, a liliaceous species not producing the analogue, Asparagus officinalis, activated it. 5. Of the other proline analogues investigated, only 3,4-dehydro-DL-proline and L-thiazolidine-4-carboxylic acid were active with the enzyme preparation from Pha8eolus. 6. pH optima of 7*9 and 8-4 were established for the enzymes from Phaseolu,s and Polygonatum respectively. 7. The Phaseolus enzyme was specific for ATP and PPi. Mn2+ partially replaced the requirement for Mg2+ as cofactor. Preincubation with p-chloromercuribenzoate at a concentration of 0 5mM or higher produced over 99% inhibition ofthePhaseolus enzyme. One-half the enzymic activity was destroyed by preheating for 5min. at 620 in tris-hydrochloric acid buffer, pH7.9. 8. All experimental evidence supports the hypothesis that azetidine-2-carboxylic acid and proline are activated by the same enzyme in Phaseolus preparations, whereas the analogue was inactive in all Polygonatum preparations. The possible nature of this different substrate behaviour is discussed.