The statistical properties of two methods of estimating gain scores for groups in quasi-experiments are compared: (1) gains in scores standardized separately for each group; and (2) analysis of covariance with estimated true pretest scores. The fan spread hypothesis is assumed for groups but not necessarily assumed for members of the groups. Sample standard deviations provide a biased estimate of treatment effects when the first procedure is used with small samples and poorly correlated measures. Under a nonlinear model of within-group growth, the true score analysis of covariance underadjusts for initial group differences. Under the assumption of linear within-group growth and with large samples, both methods estimate the desired effect with equal precision. When samples are small, neither of the two methods are appropriate for the nonlinear model of within-group growth. (Author/CTM)