The prevalence of obstructive sleep apnea (OSA) is high worldwide. This study aimed to quantify the relationship between the incidence of OSA and sleep patterns and genetic susceptibility. A total of 355,133 white British participants enrolled in the UK Biobank between 2006 and 2010 with follow-up data until September 2021 were recruited. We evaluated sleep patterns using a customized sleep scoring method based on the low-risk sleep phenotype, defined as follows: morning chronotype, 7-8 hours of sleep per day, never/rarely experience insomnia, no snoring, no frequent daytime sleepiness, never/rarely nap, and easily getting up early. The polygenic risk score was calculated to assess genetic susceptibility to OSA. Cox proportional hazard models were used to evaluate the associations between OSA and sleep patterns and genetic susceptibility. During a mean follow-up of 12.57 years, 4618 participants were diagnosed with OSA (age: 56.83 ± 7.69 years, women: 31.3%). Compared with those with a poor sleep pattern, participants with a normal (HR: 0.42, 95% CI: 0.38-0.46), ideal (HR: 0.21, 95% CI: 0.19-0.24), or optimal (HR: 0.15, 95% CI: 0.12-0.18) sleep pattern were significantly more likely to have OSA. The genetic susceptibility of 173,239 participants was calculated, and the results showed that poor (HR: 3.67, 95% CI: 2.95-4.57) and normal (HR: 1.89, 95% CI: 1.66-2.16) sleep patterns with high genetic susceptibility can increase the risk for OSA. This large-scale prospective study provides evidence suggesting that sleep patterns across seven low-risk sleep phenotypes may protect against OSA in individuals with varying degrees of genetic susceptibility.