New-onset diabetes (NOD) associated with antihypertensive medications is controversial. A 2007 network meta-analysis included 22 clinical trials reported through 15 SEP 06. Since then, data from 14 more trials (e.g., TRANSCEND, PRoFESS, ONTARGET, NAVIGATOR) have been published and one retracted (Kyoto), resulting in a total of 17,235 subjects with NOD (from a total of 217,820 at risk). Meta-analyses were performed using the network technique of Lumley (Stat Med. 2002;21:2313-24) and the traditional Bayesian method (WINBUGS Version 1.4.1, non-informative priors, 99,000 iterations, 1000 burn). For each therapy, the number of new cases of incident diabetes/number at risk (number of trials) were: Placebo: 4879/52,700 (18); Diuretic: 1422/28,786 (15); Beta-blocker: 3043/43,084 (16); Calcium channel blocker (CCB): 3204/47,928 (18); Angiotensin converting-enzyme (ACE)-inhibitor: 2564/40,534 (13); Angiotensin receptor blocker (ARB): 4099/44,459 (16). These numbers reflect ‘‘double-counting’’ for the 6 trials that included more than 2 arms (AASK, ALLHAT, COPE, MRC-E, ONTARGET, STOP-Hypertension 2). The incoherence value (u) for the network meta-analysis was 0.048. The results were: