In a multi-stage sample design, primary sampling units (PSUs) are generally selected using systematic probability proportional to size (PPS) sampling. Certainty PSUs are usually identified up front iteratively and include those PSUs whose measure of size (MOS) exceeds the sampling interval at each iteration. Then all identified certainty PSUs and a sample of non-certainty PSUs are selected. Sometimes instead of certainty PSUs being identified up front, a multi-hit approach is used where a systematic sampling skip interval is applied through all PSUs to identify certainty PSUs in one pass. The large PSUs with MOS greater than the skip interval can receive one or more hits and are identified as certainty PSUs. A cluster of ultimate sampling units is selected from each non-certainty PSU while the number of clusters selected from each certainty PSU is equal to the number of hits a certainty PSU receives. This talk will discuss alternative methods of computing selection probabilities and base weights, and will compare relative efficiencies of corresponding estimates when a multi-hit approach to PSU selection is used.