Ca2+ sparks constitute the fundamental units of sarcoplasmic reticulum (SR) Ca2+ release in cardiomyocytes. However, despite more than 25 years of investigation, the precise nature by which ryanodine receptors (RyRs) collaborate to generate these release events remains unclear. This challenge is related to both technical limitations in imaging RyRs and the rapid time frame in which sparks occur. Unfortunately, various imaging techniques capable of resolving RyRs, including super-resolution microscopy (dSTORM, DNA-PAINT) and electron microscopy, require fixed samples. To circumvent this limitation, we developed a transgenic mouse with photo-activated (PA) tagRFP targeted to RyR2. This approach allows correlative pairing of RyR localization, determined by PA Localization Microscopy (PALM), and Ca2+ sparks detected by high-speed imaging with a highly inclined light sheet (HILO). Ca2+ spark recordings showed that a subset of events exhibited slow kinetics, with protracted rise times and durations. Subtracting estimated Ca2+ diffusion revealed that prolonged Ca2+ sparks exhibited multiple distinct releases, numbering between 2 and 8 events. Notably, consecutive releases were associated with displacement of the spark centroid. Paired imaging of RyRs confirmed that these “travelling sparks” moved between nearby RyR clusters, with some sparks exhibiting displacement as far as 500 nm along z-lines. Importantly, spark propagation often proceeded between clusters that were not within the closest proximity. Treatment with isoproterenol exaggerated this phenomenon, as a larger fraction of travelling sparks was observed which included as many as 12 distinct release sites. These data suggest that participation of discrete RyR clusters in Ca2+ spark generation is dependent not only on RyR cluster position, but also other factors such as local post-translational modifications which are critically altered during β-adrenergic stimulation.