The Centers for Disease Control and Prevention (CDC) recommend a two-test approach for the serodiagnosis of Lyme disease (LD), with EIA testing followed by Western immunoblotting (WB) of EIA-equivocal and -positive specimens. This approach was compared with a simplified two-test approach (WB of EIA equivocals only) and WB alone for early LD. Case-patients with erythema migrans (EM) rash >/=5 cm were recruited from three primary-care practices in LD-endemic areas to provide acute- (S1) and convalescent-phase serum specimens (S2). The simplified approach had the highest sensitivity when either S1 or S2 samples were tested, nearly doubling when S2 were tested, while decreasing slightly for the other two approaches. Accordingly, the simplified approach had the lowest negative likelihood ratio for either S1 or S2. For early LD with EM, the simplified approach performed well and was less costly than the other testing approaches since less WB is required.