Background: Severe mental illness (SMI) among imprisoned individuals is a global health concern. Quantitative research indicates on average mental health symptom improvements during imprisonment, however, it cannot reflect multifaceted factors influencing the course of SMI. This study aimed to explore the subjective course of SMI during imprisonment and to identify influencing factors. Methods: The study has a 3-year-follow-up design of imprisoned individuals in Chile. We conducted semi-structured interviews with 10 men and 9 women at follow-up who had either major depression or psychosis (severe mental illnesses) at baseline. We included individuals who deteriorated, remained stable or improved their mental health according to quantitative assessments. We explored the subjective course of their mental health condition during the follow-up period. Qualitative data was transcribed and coded using NVivo Software for quantitative content analysis. Qualitative data was also manually coded and was subsequently analyzed using the thematic analysis method with an inductive approach. We developed the final themes using the results of this analysis in combination with the inclusion criteria. Results: The mental health of 10 individuals subjectively improved, 6 experienced deteriorations, and 3 did not perceive any change. Good infrastructure, structure/occupation, intrapersonal resources (will to change, spirituality) and supportive relationships were identified as factors improving mental health. Factors deteriorating mental health were identified as bad infrastructure (e.g., no running water and bad housing conditions) and crowding, lack of psychological treatment, exposure to violence, interpersonal stress (conflictive relationships and separation from family), perceived injustice through sentencing, intrapersonal stressors and previous medical conditions. Discussion: SMI in prison can improve in a supportive environment under certain