Diabetes is associated with an increased risk of heart failure, cardiac arrhythmias and sudden cardiac death. We previously showed that ROCK2 expression is elevated in diabetic rat hearts, and that ROCK inhibition acutely improves their contractile function. In the present study we investigated whether inhibition of ROCK or partial deletion of ROCK2 improves impaired Ca2+ handling in the diabetic heart. Contractile properties and Ca2+ transients were measured before and after treatment with the ROCK inhibitor Y-27632 (1 μM) in fluo-4-loaded cardiomyocytes isolated from streptozotocin (STZ)-diabetic or non-diabetic rats. Cardiac function was determined in vivo, and contractile properties and Ca2+ transients also measured in cardiomyocytes from non-diabetic and STZ-diabetic wild-type (WT) and ROCK2+/− mice. ROCK inhibition improved some parameters of contractile function and Ca2+ handling in cardiomyocytes from diabetic rat hearts. In addition, ROCK inhibition attenuated the diabetes-induced delayed aftercontractions (DACs) and associated irregular Ca2+ transients induced by increased Ca2+o. Although no overt cardiac dysfunction was detected in diabetic WT mice, cardiomyocytes from these mice also developed arrhythmic Ca2+ transients in response to increased Ca2+. These were attenuated in cardiomyocytes from diabetic ROCK2+/− mice, in association with decreased diastolic Ca2+ leak and with reduction of the diabetes-induced increased phosphorylation of both CaMKII and the ryanodine receptor (RyR). These data suggest that ROCK2 contributes to diabetes-induced impaired cardiac Ca2+ homeostasis, at least in part by promoting CaMKII-mediated phosphorylation of RyR. This may have important clinical implications for the treatment of the increased incidence of dysrhythmias in diabetes. •ROCK inhibition improved diabetes-induced Ca2+ transient and contraction aberrations.•In diabetes, ROCK2 inhibition or partial deletion attenuated delayed aftercontractions and Ca2+ waves.•Diastolic Ca2+ leak and elevated RyR2 phosphorylation were prevented by partial ROCK2 deletion.•This was associated with attenuation of CAMKII phosphorylation and activation.