Four different types of contact tracing for partners of Chlamydia trachomatis infected patients were compared with respect to reinfections during two 15-month periods in 1979/80 and in 1983/84. Between the two study periods a gradual improvement in contact tracing occurred. No reduction in overall reinfections was observed in the second period, reinfections occurring within 12 weeks were similar in the two periods. Treatment of the regular partner (by giving the woman a prescription for her partner) without examination or further partner follow-up resulted in few reinfections. A system of partner referral without verification that partners actually appeared for examination resulted in significantly more reinfections by the regular partner than a system with non-mandatory verification: 60 of 997 women (6%) compared with 6 of 645 (1%). The addition of C. trachomatis infection to the existing STD legislation resulted in fewer identified partners refusing examination than previously, one of 254 men (0.4%) compared with 40 of 223 (18%) before legal inclusion.