Streptomyces natalensis is an actinomycete which produces the antifimgal agent pimaric in (also known as natamycin) as a second arymetabolite. This research involves an investigation into the growth and morphology of Streptomyces natalensis in submerged culture with a view to optimising both characteristics. Four maintenance and cultivation techniques were investigated, involving the use of different inoculum sizes and different inoculum media, in a preliminary attempt to optimise grow than dpimaric in production. Analgorithm was developed and validated to characterise the morphology of the organism, (in terms of hyphalelements, clumps orpellets) using an Optimas Image Analyser (as emi-automated image analysis system). Further, this system and a Lasentec Focused Beam Reflectance Measurement (FBRM) probe were comparatively evaluated in terms of their potential to develop a morphologicalpro file of the organism. On the basis of the work performed it was determined that maintenance and cultivation techniques significantly affected growth and secondary metabolite production. It was concluded that YED was the preferred in oculummedium in term so fbiomass and subsequent pimaric in production levels. Two fermentation media , SPG and phosphate- limited medium, were considered suitable for production of the secondary metabolite. Recorded pimaric in levels were, however, significantly lower than those obtained by other researchers. Cultivation of the organism in YEME medium, using a sporeinoculum, resulted in high levels of biomass (~ 5 g/L). Validation of the image analysis technique included an investigation of these insitivity of the re su lts to (i) the level of microscopic magnification employed, (ii) sample dilution (iii) the number of entities counted. Ensuring that 10-15 % of the to talarea available in the region of interest was occupied by entities of interest, as used by Treskatis et al. (1997), it was determined that x lO (as opposed to x 4 magnification) gave a more accurate representation of the morphology present. Also, having investigated samples of up to 1000 entities, it was concluded that measurement of 600 entities was sufficient to accurately portray the morphological state of the organism. Volume-based analysis of the morphological data (as opposed to number-based) was found to be beneficial in terms of providing a more accurate portrayal of the existing morphology, using the image analyser. Hyphal contribution, on this basis, was discovered to be negligible. During the early stages of a batch fermentation, numbers and sizes of both ‘larger’ and ‘smaller’ entities increase, followed by a subsequent reduction (and corresponding increase) in the former and latter, respectively, upon entry of the organism into the stationary phase. Limited trials performed using the FBRM probe suggested the potential of this system for on-line morphological measurements of fermentation organisms. However this work also highlighted the inherent difficulty of drawing direct comparisons between data collected using an image analyzer and those resulting from the FBRM.