The influence of three bacteria strains (Helgo 21, NCMB 308 and NCMB 13) on the life cycle of Tisbe holothuriae Humes 1957 was investigated under constant experimental conditions (19-20~ and 38%o S). For each of the first six experiments, females (F1), whose life history was followed, were obtained from a common mother (F0). Experiment n ~ 7 was carried out with adult females (F2) obtained during exp. 4. Females were bred in 50-ml dishes and transferred to a fresh container as soon as they produced a new egg-sac. Offspring (F2: exp. 1-6; F3: exp. 7) from successive egg-layings were counted as soon as they moulted into adults. Larval mortality was estimated by enumeration of the adults issuing from a known number of nauplii that had been fed different bacteria strains. Some difficulties arose in obtaining adults with NCMB 13: it appeared to be necessary to provide this first generation with a small quantity of TetraMin (commercial fish food) in order to initiate reproduction and obtain a second generation. However, in exp. 4, F2 adult females fed NCMB 13 became ovigerous and fertile without needing TetraMin. This result is interpreted as denoting an indispensable period of adaptation to the diet, which probably requires important changes at the enzymatic level. One experiment was carried out with Helgo 21; another one with NCMB 308; two with NCMB 13 (with a unique additional meal of TetraMin to allow reproduction), two with NCMB 13 during larval development, then only TetraMin during adult stage, and one with NCMB 13 alone (with females issued from exp. 4). Data obtained regarding longevity, duration of reproduction period, number of egg-sacs and number of adult descendants show an important variability between bacteria strains. Lower results were almost always obtained with Helgo 21 and NCMB 308. An interesting result lies in the maximum number of egg-sacs: low with Helgo 21 and NCMB 308 (5), but very ample with NCMB 13 (11 in one experiment). In experiment 7, carried out with animals bearing a large coefficient of inbreeding (0.375), the results concerning the maximum number of egg-sacs are good (10) as are the mean number of descendants per female (294: the highest number of the seven experiments). These results indicate that genetic factors involved in the reduction of descendants in ageing laboratory cultures can be suppressed to a certain extent by nutritional factors. The poor results obtained with Helgo 21 and NCMB 308 alone must be considered with caution: after a period of adaptation with an alimentary complement, these strains could perhaps permit an interesting production of copepods.