Biological invasions frequently begin with small populations of founding individuals which results in a genetic bottleneck.  The success or failure of introductions is often attributed to the amount and type of genetic variation that persists following introduction.  Although molecular markers are often used to assess genetic variability, they are not always indicative of variability in traits related to fitness.  To effectively use marker variation to estimate morphological and phonological variation, it is necessary to compare molecular variation to quantitative genetic variation with methods like Fst-Qst comparisons.  I have examined both neutral molecular and quantitative genetic variation in both native and exotic populations of the annual plant, Impatiens capensis, which escaped from garden cultivation into two watersheds in England after introduction from its native North American range.  By comparing molecular, morphological, and phenological variation within and between both native and escaped ornamental populations, I found evidence suggestive of a genetic bottleneck in molecular markers but not in morphological or phonological traits.  Fst estimated from AFLP markers was only 0.17 between six exotic populations in England, whereas it was 0.37 between 10 native populations in Southeastern New England.  Qst was 0.35 for seed dormancy in both the native and exotic range, whereas Qst for a summary principal component of aboveground elongation traits was 0.42 in the exotic range and only 0.25 in the native range.  These results suggest that at least one set of widely used molecular markers are not reliable indicators of variation in biologically meaningful traits in some biological invasions.