People with a diagnosis of schizophrenia often have poor sleep, even when their psychotic symptoms are relatively well managed. This includes insomnia, sleep apnoea, hypersomnia, and irregular or non-24 h sleep–wake timing. Improving sleep would better support recovery, yet few evidence-based sleep treatments are offered to this group. This paper presents a mixed methods feasibility and acceptability study of Light–Dark and Activity Rhythm Therapy (L-DART). L-DART is delivered by an occupational therapist over 12 weeks. It is highly personalisable to sleep phenotypes and circumstances. Ten participants with schizophrenia spectrum diagnoses and sleep problems received L-DART; their sleep problems and therapy goals were diverse. We measured recruitment, attrition, session attendance, and adverse effects, and qualitatively explored acceptability, engagement, component delivery, adherence, activity patterns, dynamic light exposure, self-reported sleep, wellbeing, and functioning. Recruitment was ahead of target, there was no attrition, and all participants received the minimum ‘dose’ of sessions. Acceptability assessed via qualitative reports and satisfaction ratings was good. Adherence to individual intervention components varied, despite high participant motivation. All made some potentially helpful behaviour changes. Positive sleep and functioning outcomes were reported qualitatively as well as in outcome measures. The findings above support testing the intervention in a larger randomised trial ISRCTN11998005.