Particle deposits in an inertial impactor stage can be classified as primary or secondary. The primary deposits are directly downstream of each nozzle. Secondary deposits can be in the form of a halo around the primary deposit or of straight-lines between adjacent nozzles. Certainly, these additional secondary deposits lead to a deviation from the theoretical efficiency curve in which only a single primary deposit existed and was considered. The jet-to-jet interaction in multi-nozzle inertial impactors and the formation of secondary deposits has been analyzed and quantified as a function of the distance between nozzles and the Reynolds number. This work has been studied experimentally (Part I) and numerically (Part II), using CFD (Computational Fluid Dynamics) analysis to provide insight into the physical mechanisms for the formation of the secondary deposits and to support the experimental results. This paper provides detailed experimental impactor collection efficiency curves for impactor plates with 3 nozzles spaced in the range of 2.5W to 8W, where W is the nozzle diameter, and Reynolds numbers between 465 and 2210, which quantify the effects of the jet-to-jet interactions and the magnitude of the primary and secondary deposits. From the experimental results it may be concluded that the jet-to-jet interactions can produce four non-ideal effects: premature primary deposits with a half-moon shape, straight-line deposits between nozzles, nozzle plate deposits between nozzle exits (losses), and dispersion of the halo deposits. The secondary deposits and premature primary deposits reduce the slope of the efficiency curve, and/or displace the root Stk(50) towards smaller values and/or generate tails at the low end. All of which result in the collection of small particles that otherwise would be collected at another stage with a lower cut point. With respect to the design criteria, to avoid the formation of the additional deposits generated by jet-to-jet interactions (premature primary deposits, straight-line and nozzle plate deposits), the nozzle-to-nozzle spacings have to be larger than 4W and/or the impactor should work at low Re (in this work Re <= 465), even at 2.5W spacing.