When Missouri Department of Transportation (MoDOT) implemented design–build (DB) contracting, it revised its quality assurance programme and shifted most of the compaction testing to the design-builder. As a result, fewer compaction tests were performed by state personnel and the need for speedy quality control testing by the agency to facilitate construction production disappeared. This paper reports the results of a study conducted by the department to evaluate three alternatives to the nuclear density gauge (NDG) using life cycle cost analysis and cost index number theory. The study’s objective was to investigate alternative soil compaction test devices and provide input to a decision regarding whether or not MoDOT should retain or replace the NDG. Despite the NDG successful track record, the ease of employment and speed with which the compaction results are delivered comes with a price in terms of life cycle costs. The NDG is regulated by the Nuclear Regulatory Commission and entails an onerous, ongoing administrative workload to permit its continued use. The NDG also incurs additional certification, storage and disposal costs, not found in non-nuclear compaction testing alternatives. This paper reports the results of a life cycle cost analysis of NDG and three alternatives: dynamic cone penetrometer (DCP), electrical density gauge and the sand cone (SC). The study finds that the SC and DCP are the most cost-effective but are the least cost-effective when measured on a basis of timely results. Thus, the NDG replacement/retention decision becomes one of how fast are compaction tests required by the agency. Since MoDOT has adopted contractor acceptance testing in its DB program, it now only conducts verification testing of contractor test results. Thus, the paper recommends that the NDG be replaced.