The rate and extent of rumen fermentation of different starch sources can be very different, depending on the origin of the starch, but more importantly on the technological treatment of the starchy feed ingredients. Therefore, feeding different starchy feed ingredients can contribute in a very different way to the total volatile fatty acid (VFA) production in the rumen, the proportion of each VFA, and the non-glucogenic to glucogenic ratio (NGR). For 14 different starchy feed ingredients, the in vitro fermentation characteristics were determined using the gas production technique. Highest rate and extent of gas production was seen for the popped feedstuffs, while lowest was seen for the native starch sources maize and potato. This was also reflected in the rate of production of individual and total VFA. However, after 12 h of fermentation, differences in VFA content decreased and VFA production reflected the total fermentation of the organic matter. It proved that for the determined incubation periods (4, 8 and 12 h), there was a negative linear relationship (R2 = 0.33–0.79) between NGR and the amount of gas produced. There was also a negative linear relationship (R2 = 0.75) between the synthesized amount of microbial protein and the rate of fermentation at the incubation period at which the substrate was just exhausted (tRmax2). This shows that fast fermenting substrates resulted in a higher amount of microbial protein than slowly fermenting substrates. Consequently, there was also a negative linear relationship (R2 = 0.64) between the amount of microbial protein and NGR at tRmax2. It is concluded that fast fermenting starchy feedstuffs resulted in higher amounts of microbial protein in the rumen and a more glucogenic fermentation pattern, higher values of propionic acid and lower values of acetic acid and butyric acid.