**Background:** Patients with schizophrenia and other severe mental illnesses (SMI) have a life expectancy 15–20 years shorter than the general population, largely due to preventable cardiovascular disease. Smoking is highly prevalent in this population, yet few studies have examined smoking habits and attitudes among patients in residential rehabilitation facilities, particularly in rural or insular regions of Greece. This study aimed to explore smoking-related attitudes and behaviors in patients with SMI living in residential facilities on the island of Corfu, Northwest Greece.
**Methods:** This was a cross-sectional, quantitative, descriptive, correlation study conducted in nine residential rehabilitation facilities run by the General Hospital of Corfu. Of 153 total residents, 23 met exclusion criteria (10 with comorbid mental retardation, 13 with severe cognitive disturbance). Of the remaining 130, 11 declined consent and 16 did not return the questionnaire, yielding a final sample of 103 participants. All participants had a chronic SMI diagnosis (ICD-10 F20–29 or F31), were in a stable phase of illness, and provided written informed consent. Data were collected using a 33-item questionnaire (Kourakos et al., 2015) covering sociodemographics, clinical characteristics, smoking habits, and attitudes toward smoking (10 items on health, 11 on quitting difficulty, 12 on facility smoking). Factor analysis yielded a 3-factor structure explaining 54.12% of variance (Factor 1: difficulty quitting, 25.44%; Factor 2: avoiding smoking in facilities, 16.33%; Factor 3: confidence in a non-smoking reference person, 12.35%). Cronbach's alpha was 0.611 overall (Factor 1: 0.668, Factor 2: 0.754, Factor 3: 0.926). Statistical analyses included t-tests, Kruskal–Wallis tests, post hoc Bonferroni corrections, Spearman correlations, and chi-square tests, with p < 0.05 considered significant.
**Key Results:** Of the 103 participants, 63.1% were male, 60.2% were aged 41–60 years, 76.7% were unmarried, and 54.4% had only primary education. The most common diagnosis was schizophrenia (71.8%), followed by other psychotic disorders (9.7%) and bipolar disorder (8.7%); 91.3% were on antipsychotics. Nearly half (49.5%) had lived in the facility for 6–10 years. A total of 87.4% reported having smoked in the past, and 68.9% (71 patients) were current regular smokers. Mean age of smoking initiation was 21.6 years (SD 8.2), mean smoking duration was 29.3 years (SD 12.2), and mean daily consumption was 18.5 cigarettes (SD 13.8), mostly filtered (85.9%). Reasons for starting included friends (33.8%), stress/personal issues (33.8%), and curiosity (23.9%). Only 45.6% believed smoking causes severe harm to health, and 50.7% reported receiving physician advice to quit. Among current smokers, 64.8% had tried to quit, but 57.7% felt they would need help to succeed. Patients agreed that quitting is difficult (mean 4.24/5 on Likert scale, SD 1.21). Statistically significant findings included: age correlated negatively with current smoking (r = −0.239, p < 0.05) and positively with years of smoking (r = 0.717, p < 0.01); primary-education-only patients had more smoking years than secondary-education patients (p = 0.044); patients not on antidepressants reported more smoking years [t(65) = 2.255, p = 0.027]; all patients on antiepileptics were current smokers (χ² = 6.021, p < 0.01); female patients avoided smoking in the facility more than males [t(101) = −2.73, p = 0.007]; patients on mood stabilizers preferred not to smoke in facilities [t(102) = 3.128, p = 0.002]; and patients raised in semi-urban areas had greater difficulty quitting than those from urban (p = 0.095) or rural (p = 0.062) areas. Longer facility stay correlated with current smoking (r = 0.375, p < 0.01), quit attempts (r = 0.317, p < 0.01), physician advice (r = 0.328, p < 0.01), belief that smoking is harmful (r = 0.257, p < 0.05), and belief that smoking should be avoided in facilities (r = 0.365, p < 0.01). Notably, 70.9% of patients reported seeing staff smoke at work, mostly outside (68.5%) or in a smoking lounge (45.2%). Patients agreed that staff should set a good example by not smoking (mean 3.73/5, SD 0.83) and should encourage smoking cessation (mean 3.71/5, SD 0.91).
**Clinical Implications:** This study demonstrates that smoking rates are extremely high (68.9%) among SMI patients in residential facilities in insular Greece, with most patients smoking heavily for decades. Importantly, a majority have tried to quit and recognize the need for help, yet only half receive physician advice. The finding that longer-stay patients are more aware of smoking's harms and more motivated to quit suggests that residential facilities can be opportune settings for cessation interventions. The high prevalence of staff smoking observed by patients (70.9%) and patients' desire for staff to model non-smoking behavior indicate that staff smoking policies and role-modeling are critical targets. Clinicians should not assume patients are uninterested in quitting; rather, they should proactively offer counseling, pharmacotherapy (e.g., nicotine substitutes), and structured cessation programs. Patients with cardiovascular comorbidity should be prioritized. Limitations include the modest sample size (n = 103), lack of sub-analyses by diagnosis (only 9 with bipolar disorder), reliance on self-report, and absence of current clinical status assessment. Nonetheless, the findings underscore the urgent need for smoking cessation programs tailored to the institutional and cultural context of Greek residential psychiatric facilities.