In autoimmune hepatitis (AIH), the smooth-muscle antibody is specific for polymerized actin. Detection of antiactin antibody (AAA) has been hampered by technical problems. We have investigated AAA in 30 sera from patients with liver diseases and smooth-muscle antibody. AAA was detected by indirect immunofluorescence in 1:40, 1:80, and 1:160 dilutions. Five techniques were performed using fibroblasts: with vinblastine (A); without drugs (B); with sodium citrate (C); without drugs but with heat serum inactivation (D); and with sodium citrate and heat serum inactivation (E). For comparative analysis, we considered: the total number of AAA-positive sera regardless of the dilution in which reactivity was observed, as well as in each dilution separately; and the comparison of AAA intensity between 1:40 x 1:80, 1:40 x 1:160, and 1:80 x 1:160 dilutions. AAA was more positive in techniques B, C, D, and E than in A (P < .001) in general, and in each dilution separately. AAA was more positive in technique D than in B in 1:40 (P = .0005) and 1:80 dilutions (P = .03), as well as in E than in C (P = .0001) in 1:40 dilution. Techniques B and D yielded results similar to C and E, respectively. AAA staining was significantly more intense in 1:80 and 1:160 than in 1:40 dilution in A, B, and C; it was both significantly less intense in 1:80 and 1:160 than in 1:40 dilution and in 1:80 than in 1:160 in techniques D and E. We concluded that heat inactivation increased AAA seropositivity/intensity in 1:40 and 1:80 dilutions, preventing false-negative results; actin polymerization with sodium citrate did not enhanced AAA seropositivity/intensity. The technique with vinblastine was the least effective.