The Ca2+ load in the sarcoplasmic reticulum (SR) is an important stimulator of Ca2+ release which is mediated by the ryanodine receptors (RyRs) in muscle. Two quite different mechanisms have been proposed, and there is no consensus on how the Ca2+ load in the SR alters RyR activation. The “true luminal regulation” hypothesis attributes luminal Ca2+-activation to Ca2+ regulatory sites on the luminal side of the RyR while the “feed-through” hypothesis proposes that luminal Ca2+ permeates the pore and binds to the cytoplasmic sites. This study proposes a resolution of the controversy based on measurements of luminal Ca2+ activation of isolated cardiac RyRs (RyR2 isoform) in artificial lipid bilayers. In the absence of Ca2+cyt the open probability (Po) of RyR2 had a voltage-dependent, bell-shaped dependence on [Ca2+]lum. At -40 mV (favouring Ca2+ feed through) Ca2+lum activates with a Ka = 50 μM and inhibits with a Ki = 600 μM. Ka and Ki markedly increased with membrane depolarisation. The mechanism of [Ca2+]lum activation appears to be luminal-triggered, Ca2+ feed-through. Detailed analysis of channel gating indicated that Ca2+ binding sites on both the luminal (Km =50 μM) and cytosolic (Km = 1 μM) sides of the protein mediate Ca2+lum activation. Ca2+ binding to either of these sites are sufficient to open RyR2. In the virtual absence of Ca2+cyt (i.e. resting [Ca2+]) the predominant opening mechanism is the luminal site which, when bound to Ca2+ opens the channel briefly (Po ∼0.3%, to ∼1 ms). Ca2+ feed-through from the luminal to cytoplasmic site prolongs channel openings (Po ∼10%, to ∼10-50 ms). In this way, Ca2+ feed-through can produce over 90% of RyR activation yet it is completely reliant on the action of Ca2+ at a luminal facing site.