Defining clear, quantifiable, and defensible engineering performance criteria for final cover systems is a persistent challenge facing landfill, remediation, and other design teams implementing geosynthetic caps in geoenvironmental applications. Existing deterministic and rule-of-thumb performance analyses are inadequate because it is impossible to compare their results to expectations or requirements for cover system behavior over their design life. This paper presents a probabilistic methodology that has been applied to establish and evaluate performance criteria for final cover settlement and drainage. This methodology considers the interaction between the long-term mechanical behavior of geosynthetics and the heterogeneous compressibility of subgrades. The results show how engineers can connect performance analyses with design objectives within a quantitative framework using these tools. An example remediation final cover design is provided to illustrate the implementation of the framework. This paper applied a probabilistic settlement model to a typical landfill design following prescriptive final cover grading design guidelines. The resulting differential settlement and percent inundated area were computed for several probabilistic simulations of post-closure settlement. Because the prescriptive design is expected to provide an acceptable level of performance, the resulting probabilistic measures of performance (inundated area, 1-ft inundated area) are taken to represent the implied probabilistic design criteria expected for these designs. Considering the purpose of the design criteria with respect to final cover performance, the 1-ft inundated area seems to provide a more practical measure of cover performance and thus should be used to assess final cover performance. Several simulations of the settlement of the prescriptive final cover plateau design were conducted. In order to assess the impact of the probabilistic settlement model's input parameters, several different volume loss model k parameters were analyzed. Through a comparison of the typical landfill's performance with field measurements taken on an existing landfill, a k value of 0.5 was found to compare best to field measurements. The design mean 1-ft inundated area criterion selected based on the 40 realizations simulated is 0.2 %. The cumulative histogram of 1-ft inundated area for multiple realizations of design cases should also be compared to the cumulative histogram generated for the prescriptive design. Design case histograms plotting generally to the left of the prescriptive design histogram indicate better performance (less slope reversals) than the prescriptive design. It is expected that the method used to develop this criterion, when applied to probabilistic designs of final cover systems, will provide levels of final cover system reliability and performance comparable to current state-of-practice design guidance. Designs providing lower values of 1-ft percent inundated area compared to appropriate reference design cases will provide better performance than existing practice guidelines suggest. Accordingly, final cover designs can be iteratively analyzed and their performance evaluated relative to this criterion according to the demonstrated procedures. An example cumulative histogram showing a single design iteration was presented to illustrate this approach.