Riesling musts, with or without sulfur dioxide added, were fermented either with or without the addition of yeast. Uninoculated fermentations took much longer to finish than inoculated musts. There were no significant differences in growth of non‐Saccharomyces yeasts in uninoculated musts with less than 50 mg l−1 SO2 added. The starter culture was completely dominant over indigenous Saccharomyces cerevisiae and strongly inhibitory to non‐Saccharomyces. Alcohol and acetaldehyde were greater in the inoculated treatments ; titratable acidity and acetic acid were greater in the uninoculated fermentations. There were no statistically significant differences among any treatments in final pH, ammonia content, or colour (A420). Uninoculated fermentations had higher sensory scores (P > 0·95) for ‘spicy’, ‘apple’, ‘melon’, ‘pear’, and ‘H2S’, while inoculated wines had higher scores (P > 0·95) for ‘paper’, ‘oxidized’, and ‘sweaty’. Sulfite treatment produced an assortment of significant sensory differences in the finished uninoculated wines, but in inoculated wines the additions of SO2 to the must had no significant effect on indigenous yeast populations or on flavour.