To date, most of the evidence suggests that smoking is negatively associated with testing positive for SARS-CoV-2. However, evidence has several methodological limitations. Using an outpatient sample population, we analyzed the association of testing positive for SARS-CoV-2 and smoking considering comorbidities, socioeconomic and demographic factors. Baseline data were obtained from a cohort during the first wave of the pandemic in Geneva, Switzerland (March-April 2020). RT-PCR tests were carried out on individuals suspected of having SARS-CoV-2 according to the testing strategy at that time. Logistic regressions were performed to test the association of smoking and testing positive for SARS-CoV-2 and further adjusted for comorbidities, socioeconomic and demographic factors. The sample included 5,169 participants; 60% were women and the mean age was 41 years. The unadjusted OR for testing positive for SARS-CoV-2 was 0.46 (CI: 0.38-0.54). After adjustment for comorbidities, socioeconomic and demographic factors, smoking was still negatively associated with testing positive for SARS-CoV-2 (OR: 0.44; CI: 0.35-0.77). Women (OR: 0.79; CI: 0.69-0.91), higher postal income (OR: 0.97; CI: 0.95-0.99), having respiratory (OR: 0.68; CI: 0.55-0.84) and immunosuppressive disorders (OR: 0.63; CI: 0.44-0.88) also showed independent negative associations with a positive test for SARS-CoV-2. Smoking was negatively associated with a positive test for SARS-CoV-2 independently of comorbidities, socioeconomic and demographic factors. Since having respiratory or immunosuppressive conditions and being females and healthcare workers were similarly negatively associated with SARS-CoV-2 positive testing, we hypothesize that risk factor-related protective or testing behaviors could have induced a negative association with SARS-CoV-2.