Correlation among sibling marker genotypes may invalidate the results of family-based tests of association in the presence of linkage. We apply an empirical variance estimation method, which is implemented in the program package FBAT, on Caucasian families with asthma in the presence and absence of linkage and compare the results with those obtained using the TDT (TDTEX-PAIRS) on the same data sets. Our results indicate that both tests generally perform comparably in either setting.