Data were presented in a previous paper having to do with ethnic, sex, and generational differences in the flushing response to alcohol, the number of drinks required to evoke flushing, and family resemblances in flushing. Using a much largerN (1288 subjects; 245 families), the present report is of the same analyses, as well as analyses pertaining to the extent and duration of flushing, association of flushing with other symptoms, and degree to which flushing influenced drinking behavior. As in the previous paper, a distinction was made between fast (one drink or less to evoke flushing) and slow (two drinks or more to evoke flushing) flushing. Substantial ethnic differences were found in flushing. Highly significant family resemblances in flushing were also found, but these did not conform to a single-locus genetic model. Extent of flushing was significantly related to flushing type-fast vs. slow-but duration of flushing was not, nor were most other symptoms resulting from alcohol use. Fast vs. slow flushing may or may not involve differing genetic mechanisms (probably not, on the basis of data acquired and analyzed to date). However, the two flushing types appear to have different amounts of influence on drinking behavior. Regardless of flushing type, a substantial proportion of flushers does not eschew alcohol because of flushing or cease to drink once flushing occurs.