Vitamin D deficiency (VDD) may be associated with obstructive sleep apnea (OSA) presence and is more pronounced with increasing OSA severity; however, the relationship between these two entities remains unclear. This was a cross-sectional study among 262 adults with in-hospital-attended polysomnography-diagnosed OSA and no additional major comorbidities, aiming to explore possible associations between serum 25-hydroxyvitamin D [25(OH)D] levels and polysomnographic parameters. Data on demographics, medical history, anthropometric indices, and lifestyle habits were collected at enrolment. Serum 25(OH)D was evaluated using chemiluminescence, with VDD defined as 25(OH)D < 20 ng/mL. VDD was observed in 63% of the participants. Serum 25(OH)D correlated negatively with apnea–hypopnea index and other polysomnographic indices (all p < 0.05). In logistic regression analysis, adjusting for age, sex, smoking, body mass index, physical activity, dietary vitamin D intake, and season of blood sampling, serum 25(OH)D was associated with lower odds of severe OSA [odds ratio (95% confidence interval): 0.94 (0.90–0.98)]. In the same multivariate model, VDD was associated with ~threefold higher odds of severe OSA [2.75 (1.38–5.48)]. In stratified analyses, VDD predicted OSA severity in the group of participants ≥50 y [3.54 (1.29–9.68)] and among those with body mass index ≥ 30 kg/m2 [3.38 (1.52–7.52)], but not in the younger and non-obese adults. This study provides further evidence of an inverse association between vitamin D levels and OSA severity and underscores the importance of considering vitamin D status as a potential modifiable factor in the comprehensive management of OSA.