We have studied the effect of phospholipid chain length on the activity and molecular dynamics of reconstituted Ca-ATPase from skeletal sarcoplasmic reticulum (SR), using time-resolved phosphorescence anisotropy (TPA) and electron paramagnetic resonance (EPR). We used reconstituted Ca-ATPase in exogenous phosphatidylcholines with monounsaturated chains 14–24 carbons long, to determine their effects on the physical properties of the Ca-ATPase and to correlate these physical changes with changes in the ATPase activity. In agreement with previous studies, we found that the enzymatic activity was maximal with a chain length of 18 and decreased substantially with longer or shorter chains, Our TPA results show that chain lengths longer or shorter than the optimal 18 result in a significantly decreased mobility of the Ca-ATPase, indicated by higher residual anisotropy and suggesting extensive protein aggregation, Saturation-transfer EPR data obtained with a spin label bound to a different site also indicates substantial immobilization of the enzyme, supporting the TPA results. There is good agreement between the fractional inhibition of the Ca-ATPase activity and the fraction of the enzyme in large aggregates. Solubilization in the nonionic detergent C12E8 demonstrated that inhibition of enzyme activity is reversible. In contrast to the large effects on protein mobility, these changes in chain length had little or no effect on hydrocarbon chain mobility as detected by conventional EPR at different depths in the membrane. We conclude that the Ca-ATPase has an optimum lipid bilayer thickness, presumably matching the thickness of the hydrophobic transmembrane surface of the enzyme, and that deviation from this optimum thickness produces a hydrophobic mismatch that induces protein aggregation and hence Ca-ATPase inhibition. This is consistent with our proposal that protein dynamics and protein-protein interactions are of primary importance to the Ca-ATPase mechanism.