The aim was to examine the outcome of a multimodal treatment for selective mutism (SM). Seven children, aged three–five years, who were referred for SM were included. The treatment started at home and was continued at kindergarten for a maximum of six months, with predefined treatment goals in terms of speaking levels, from I (“Speaks to the therapist in a separate room with a parent present”) through to VI (“Speaks in all kindergarten settings without the therapist present”). The outcome measures were the teacher-reported School Speech Questionnaire (SSQ) and the treatment goal obtained (I–VI) six months after the onset of treatment, and the SSQ and Clinical Global Impression Scale (CGI) at one-year follow-up. Six children spoke in all kindergarten settings (VI) after a mean of 14 weeks treatment. One child, with more extensive neuro-developmental delay, spoke in some settings only (V). The mean SSQ score was 0.59 (SD = 0.51) at baseline compared with 2.68 (SD = 0.35) at the six-month evaluation and 2.26 (SD = 0.93) at one-year follow-up. The mean CGI score at baseline was 4.43 (SD = 0.79) compared with 1.14 (SD = 0.38) at follow-up. Home- and kindergarten-based treatment appears to be promising.