This research was conducted in order to enable primary producers to maximize the use of cheap roughage sources while still maintaining body weight during dry winter months when the crude protein (CP) content of roughage sources are at a minimum. The data obtained from this study will give an economic advantage when formulating supplements to be used during this time of the year. The aim of this study was to determine the optimum level of non-protein-nitrogen (NPN) and fermentable metabolizable energy (FME) to increase microbial protein synthesis, optimize rumen fermentation and increase digestibility of dry matter (DM) and neutral detergent fibre (NDF) in sheep fed on poor quality forages. A metabolic trial was conducted where intake of DM, organic matter (OM), NDF and CP was recorded; rumen volatile fatty acid (VFA) production was recorded as well as rumen pH over the different treatments. Microbial protein synthesis was determined by analysing purine derivatives in the urine. An in situ trial was also done to determine changes in ruminal digestibility of DM and NDF on different treatments. Five treatments were used. Treatment 1 consisted of NPN and FME balanced according to the NRC (2007) requirements for a 50kg whether, and served as a control. Treatment. Treatment 2 consisted 15% less NPN than control but the same amount of FME than control while treatment 3 consisted 15% more NPN than the control but the same amount of FME as the control treatment. Treatment 4 consisted of 15% less FME, but the same amount of NPN, than the control treatment, while treatment 5 consisted of 15% more FME, but the same amount of NPN than the control treatment. A 5 x 5 Latin square design was used in this study. Five Merino wethers were allowed to adapt to supplements which were infused directly into the rumen at 9:00 and 15:30 every day. After adaptation animals were placed in individual metabolic crates for three and given three day to adapt to crate environment. After the initial three days the sampling period commenced. Results obtained indicated that treatment had no effect on DM, OM, NDF and water intake but intake of CP was significantly increased for treatment 3 when compared to treatment 2. When intake of DM, OM, NDF and CP, related to metabolic bodyweight (W0.75) was calculated, treatment 5 resulted in lower intake of both water and NDF as compared to treatment 4. Differences between levels of FME and NPN in this study was insufficient to have an influence on DMD, OMD or NDFD however, CP degradability was increased for treatment 3 and treatment 5. Ruminal pH was unaffected by treatment. Increased levels of NH3-N for treatment 3 when compared to treatment 1 and 2, was observed. Both treatments 2 and 5 resulted significant decreases in rumen NH3-N. Treatments had no effect on the proportions of VFA produced or on the Acetate to Propionate produced ratio. Treatment 3 caused an improvement in CP an N balance when compared to treatment 1 and 2. Treatment 3, when compared to treatment 1 and 2, lead to an increase in N balance/kgW0.75. Treatment 5 caused a higher microbial protein synthesis in contrast to treatment 4. Results from the in situ trial showed a decreased a-value (solubility) for the NDF fraction of treatment 3 when compared to treatment 2. The rate of degradability (c) of both DM and NDF was increased for treatment 2 compared with treatment 3. The b, ED and PD values showed no response to treatment. / Dissertation (MSc Agric)--University of Pretoria, 2013. / gm2014 / Animal and Wildlife Sciences / unrestricted