["Addiction, EarlyView. ", "\nAbstract\n\nBackground and aims\nWe followed up a financial incentives trial aimed at maintaining smoking cessation postpartum, which showed some positive effects at one year, to assess sustained abstinence at 3–4 years postpartum.\n\n\nDesign, setting and participants\nSecondary analysis of longer‐term outcomes of a pragmatic, multi‐centre, three‐arm randomized controlled trial, recruiting 462 postpartum women in England, aged ≥16 years, having stopped smoking during pregnancy with financial incentives, validated as abstinent at end of pregnancy or early postpartum. Interventions: (1) usual care (UC); (2) UC plus up to £60 of voucher incentives offered to participants and £60 offered to significant‐other supporters, over 3 months postpartum, contingent upon validated abstinence (‘3‐months incentives’); or (3) UC plus ‘3‐months incentives’ plus £180 of vouchers offered to participants over 9 months postpartum (leading to a maximum of £240 of incentives for participants, plus £60 for significant others), contingent upon validated abstinence (‘12‐months incentives’). Participants who did not achieve sustained abstinence at one year were included as non‐abstainers but were not followed‐up.\n\n\nMeasurements\nPrimary outcome: biochemically validated abstinence at 3–4 years postpartum. Statistical significance was set at P < 0.017 to account for multiple comparisons.\n\n\nFindings\nAmong the full trial sample, 63% (289/462) of participants had self‐reported smoking status at 3–4 years postpartum; 19% (88/462) reported as abstinent, of whom 64% (56/88) underwent biochemical validation with none misclassified. Primary outcome abstinence was 16.4% (26/159) for 12‐months incentives, 12.3% (19/154) for 3‐months incentives and 7.4% (11/149) for UC. Adjusted odds ratios (OR) (95% confidence interval) for 12‐months vs. UC: OR 2.51 (1.18–5.31), P = 0.016, and for 3‐months vs. UC: 1.81 (0.83–3.98), P = 0.138. In a fully adjusted model, the difference between 12‐month incentives and UC was no longer statistically significant: OR 2.44 (1.11–5.36), P = 0.027, using the Bonferroni correction (P < 0.017). Bayes factors indicated that for 12‐months vs. UC there was good evidence for the alternative hypothesis and for 3‐months vs. UC there was inconclusive evidence for the null or alternative hypothesis.\n\n\nConclusions\nUp to £300 of voucher incentives offered over 12 months postpartum may be effective for maintaining smoking abstinence at 3–4 years postpartum compared with usual care. Up to £120 of incentives over 3 months postpartum does not appear to be beneficial compared with usual care.\n\n"]