The uncertainty and variability associated with soil parameters are conventionally accounted for in practice by the adoption of point estimates of parameters, with the estimated values reflecting the engineer's confidence level in the observed data. This approach is rather simplistic and falls short of providing enough information on the bulk of available data. The statistical and probabilistic method of analysis is a rational and systematic approach that recognizes the variability of soil properties and provides reliable estimates of soil parameters for design purposes. For homogeneous and slightly heterogeneous soils, the mean value of the parameters from probabilistic analysis compares favorably with single value estimates obtained from conventional soil analysis. Reliability assessment of the stability of a retaining wall located in a project area of slightly heterogeneous soils in the Niger Delta region has been carried out using rigorous analytical methods and Microsoft – Excel spreadsheet optimization. The two approaches produced fairly similar results for parameters obtained from conventional and probabilistic soil analysis. On this basis, it is therefore affirmed that parameters derived from conventional analysis of samples of homogenous and slightly heterogeneous soils are adequate for design. Hence, there is no need to embark on the rigorous and complex probabilistic analysis, particularly in projects of moderate scale. This may only be necessary in large scale projects and in sites where soils exhibit pronounced heterogeneity. The spreadsheet-based reliability analysis however exhibits versatility and is recommended as a convenient analytical tool in the stability analysis of retaining walls. Its ability to explicitly reflect the correlation, standard deviation, probability distributions and sensitivities and to automatically seek the most probable failure combination of parametric values for any case under consideration gives it an edge over other methods of analysis.