Abstract In practice in the UK, grasses and ruderal forbs recruited from the urban soil seed bank often have a detrimental impact on establishment and development of sown meadows. This study investigated the effect of graminicide, and sowing rate on establishment, survival and longer term development of sown meadow forbs. A seed mix containing 19 forb species was sown at two sowing rates in a randomized block experiment in Sheffield, UK. The two most abundant grasses recruited, Arrhenatherum elatius, and Holcus lanatus, were highly damaging to survival and development of sown forbs. Cutting the meadow to 50 mm during the first year did not alleviate grass competition. Weedy ruderal forbs had a lesser effect on sown forbs than the perennial grasses. Where a graminicide was used, sown forb density and biomass were significantly higher in the second and third year of the study, and forb richness significantly greater in all 3 years. Sown forb density was higher at the high sowing rate in all three years, and forb richness in years 1 and 2. Sowing rate did not significantly increase forb biomass in any year. In general, suppressing grass growth had a more beneficial effect on sown forb establishment and development than increasing sowing rate.