Smoking during pregnancy is an important problem in the Netherlands. We tested the effectiveness of a health counseling method by midwives using a RCT. Four provinces with 42 practices including 118 midwives were randomly assigned to the experimental or control condition. Midwives in the experimental group provided brief health counseling, self-help materials on smoking cessation during pregnancy and early postpartum, and a partner booklet. Controls received routine care. The main outcome measures were 7-day abstinence, continuous abstinence, and partner smoking at 6 weeks post-intervention (T1) and 6 weeks postpartum (T2). Multi-level analysis revealed significant differences between both conditions at T1 and T2 using intention-to-treat analysis. Nineteen percent of the experimental group reported 7-day abstinence compared to 7% of the control group at T1, and 21 and 12%, respectively, at T2. For continuous abstinence these percentages were 12% in the experimental group and 3% in the control group. The partner intervention was not successful. The intervention resulted in significant effects on smoking behavior for pregnant women, but not for partner smoking. The program realized short-term effects. An important precondition is that midwives need a proper training.