Obstructive sleep apnoea is a common yet frequently underdiagnosed condition in patients with type 2 diabetes, particularly in primary care. Early detection is important, as untreated sleep apnoea may contribute to worsened metabolic control and increased cardiovascular risk. This study evaluated 164 patients with type 2 diabetes and found that 75% had obstructive sleep apnoea, with 31% requiring treatment for moderate to severe cases. Predicting obstructive sleep apnoea for which medical treatment is indicated (i.e., moderate to severe OSA) proved challenging, as typical clinical symptoms and most other readily available clinical parameters proved to be unreliable indicators. However, central fat distribution, indicated by a higher waist-to-hip ratio (odds ratio 3.31, 95% confidence interval 1.91-6.25, p = 0.0032), and the presence of albuminuria (odds ratio 7.46, 95% confidence interval 1.99-27.89, p = 0.0244), emerged as significant predictors, with albuminuria representing a novel finding. Screening tools such as the STOP-Bang questionnaire had limited predictive accuracy. These findings highlight the importance of targeted screening in patients with type 2 diabetes, particularly those with central fat distribution or albuminuria, to reduce underdiagnosis and potentially improve treatment outcomes.