This meta-analysis and review of the literature examines the long-held belief that CPAP treatment improves glucose levels in patients with OSA. As usual in these cases, thousands of articles were initially flagged up in the search with only six RCTs (496 patients), making the final cut. Rather surprisingly it seems that there is no evidence that CPAP helps diabetic control. This is worrying as the gold standard treatment for OSA really ought to be able to help patients with the co-morbidities associated with this disease rather than just treat the AHI. Looking closer at the numbers it reveals the main problem with CPAP is compliance. In the study where patients were effectively forced to use CPAP (admitted and monitored for seven days), for almost eight hours a night the results were very good. In the other studies with poor compliance, the results were poor. Long-term compliance with CPAP is extremely low, when compliance is defined as patient device usage of four hours a night 70% of the time (approximately three hours a night). Long-term compliance over 7.5 hours 100% of the time is required to deal with most of the co-morbidities of OSA (especially reducing road traffic accidents), but this is very rarely achievable. There are very few long-term CPAP studies, and there are ongoing questions about the ‘gold standard’ status of CPAP as a result. This study adds significantly to the debate.